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
BMC Med Res MethodolBMC Medical Research Methodology1471-2288BioMed Central London 1471-2288-5-241609296010.1186/1471-2288-5-24Research ArticleWhich resources should be used to identify RCT/CCTs for systematic reviews: a systematic review Crumley Ellen T [email protected] Natasha [email protected] Kristie [email protected] Terry P [email protected] Lisa [email protected] HealthInfo & Searching Practice Inc., Edmonton, Canada2 Department of Medicine, Division of Nephrology, University of Alberta, 4058 Research Transition Facility, Edmonton, Alberta T6G 2E1, Canada3 Department of Pediatrics, Complementary and Alternative Research and Education (CARE) Program, University of Alberta, 4047 Research Transition Facility, Edmonton, Alberta T6G 2E1, Canada4 Department of Pediatrics, Alberta Research Centre for Child Health Evidence (ARCHE), University of Alberta, 4th Floor Aberhart Centre One, 11402 University Avenue, Edmonton, Alberta T6G 2J3, Canada2005 10 8 2005 5 24 24 23 2 2005 10 8 2005 Copyright © 2005 Crumley et al; licensee BioMed Central Ltd.2005Crumley 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
Systematic reviewers seek to comprehensively search for relevant studies and summarize these to present the most valid estimate of intervention effectiveness. The more resources searched, the higher the yield, and thus time and costs required to conduct a systematic review. While there is an abundance of evidence to suggest how extensive a search for randomized controlled trials (RCTs) should be, it is neither conclusive nor consistent. This systematic review was conducted in order to assess the value of different resources to identify trials for inclusion in systematic reviews.
Methods
Seven electronic databases, four journals and Cochrane Colloquia were searched. Key authors were contacted and references of relevant articles screened. Included studies compared two or more sources to find RCTs or controlled clinical trials (CCTs). A checklist was developed and applied to assess quality of reporting. Data were extracted by one reviewer and checked by a second. Medians and ranges for precision and recall were calculated; results were grouped by comparison. Meta-analysis was not performed due to large heterogeneity. Subgroup analyses were conducted for: search strategy (Cochrane, Simple, Complex, Index), expertise of the searcher (Cochrane, librarian, non-librarian), and study design (RCT and CCT).
Results
Sixty-four studies representing 13 electronic databases met inclusion criteria. The most common comparisons were MEDLINE vs. handsearching (n = 23), MEDLINE vs. MEDLINE+handsearching (n = 13), and MEDLINE vs. reference standard (n = 13). Quality was low, particularly for the reporting of study selection methodology. Overall, recall and precision varied substantially by comparison and ranged from 0 to 100% and 0 to 99%, respectively. The trial registries performed the best with median recall of 89% (range 84, 95) and median precision of 96.5% (96, 97), although these results are based on a small number of studies. Inadequate or inappropriate indexing was the reason most cited for missing studies. Complex and Cochrane search strategies (SS) performed better than Simple SS.
Conclusion
Multiple-source comprehensive searches are necessary to identify all RCTs for a systematic review, although indexing needs to be improved. Although trial registries demonstrated the highest recall and precision, the Cochrane SS or a Complex SS in consultation with a librarian are recommended. Continued efforts to develop CENTRAL should be supported.
==== Body
Background
The aim of systematic reviews is to present the most valid estimate of the effectiveness of the intervention in question. To do so, the identification of relevant studies must be comprehensive and unbiased. Systematic reviews usually include a comprehensive summary of data from both randomized (RCT) and controlled trials (CCT), although other study designs are sometimes incorporated. There is an ongoing debate about the number and type of resources that should be used to identify trials for systematic reviews [1-3]. These resources include electronic databases, the Internet, handsearching, checking relevant article references, and personal communication with experts in the field. Reviewers are encouraged to search numerous resources in order to identify as many relevant studies as possible without systematically introducing bias [4]. However, searching more resources typically results in a higher yield; thus, more time and resources are required to conduct the review [5]. Consequently, determining the relative value of different sources of trials is critical to enhance the efficiency of systematic reviews, while maintaining their validity.
The Cochrane Collaboration Reviewers' Handbook notes that MEDLINE, EMBASE, and CENTRAL are the three electronic bibliographic databases generally considered as the richest sources of trials [6]. The Cochrane Collaboration maintains that handsearching is vital to the credibility and success of systematic reviews [7]. Hopewell et al. conducted a systematic review of studies that compared handsearching to searching an electronic database for RCTs [8]. In the 34 included studies, they found that complex searches of electronic databases recalled 65% of relevant RCTs; the other 35% were retrieved in other ways including handsearching. They concluded that "handsearching still has a valuable role to play in identifying reports of randomized controlled trials for inclusion in systematic reviews of health care interventions, particularly in identifying trials reported as abstracts, letters and those published in languages other than English, together with all the reports published in journals not indexed in electronic databases" [8].
Our research question was: Does resource-specific searching retrieve RCT/CCTs for systematic reviews with the same recall and precision as those searches which combine two or more distinct resources? Our primary goal was to identify and quantitatively review studies comparing two or more different resources (e.g., databases, Internet, handsearching) used to identify RCTs and CCTs for systematic reviews. Specifically, we were interested in determining the value (in terms of identifying unique citations) of searching key resources (e.g., EMBASE, CENTRAL, PsycINFO, handsearching) in addition to the key resource MEDLINE.
Methods
Search strategy
Seven electronic databases (MEDLINE, EMBASE, CINAHL, ERIC, PsycINFO, Web of Science, Cochrane Library) were searched from their inception to April 2004. The MEDLINE search strategy was tailored as necessary for each database (Appendix 1). Four journals were handsearched from 1990 to 2004: Health Information & Libraries Journal (Health Libraries Review), Hypothesis, Journal of the Medical Library Association (Bulletin of the Medical Library Association), Medical Reference Services Quarterly. All abstracts presented at Cochrane Colloquia (1993–2003) were handsearched. In addition, key authors were contacted via email and references of relevant articles were screened. The searches were not limited by language or date of publication. Searches are available upon request from the corresponding author.
Study selection
Two reviewers independently screened the yield from the searches to identify potentially relevant studies. The full text of these studies was obtained and two reviewers independently applied inclusion/exclusion criteria using a standard form. Any differences were resolved through discussion. Studies were included if they compared two or more sources to find RCTs or CCTs (e.g., one or more resources compared against a "gold standard"; handsearch versus MEDLINE; and, EMBASE versus MEDLINE). Inclusion was not limited by the topic/content area in the individual studies.
A study was excluded if: it only compared different search strategies within the same database; it only included non-randomized trials; or, if the resource is not currently accessible. If authors searched for all study designs including trials, it was included only if data were reported separately for RCTs or CCTs.
RCTs were defined as an experiment in which eligible patients are assigned to two or more study groups using an appropriate method of randomization [9]. CCTs were defined similarly except that the method of allocation is not necessarily random. When authors did not provide definitions, we accepted their classification/indication of study design.
Assessment of quality
We developed a checklist to assess the quality of reporting of the included studies. The quality items were chosen based on threats to the validity of comparative studies that have been empirically supported in the literature [10]. The items assessed reporting in four key areas:
• Was there an adequate description of what the search was attempting to identify (e.g., type of studies, content area, standard inclusion/exclusion criteria);
• Was there an adequate description of the methods used to search (e.g., resource(s), words/subject headings used, time period covered, date of search);
• Was there an adequate description of the reference standard (e.g., how many references) and how it was derived (e.g., sources searched and methods used); and,
• Was bias avoided in the selection of relevant studies (e.g., was there an independent assessment of studies for inclusion by more than one researcher).
Two reviewers independently applied the checklist to the included studies. Discrepancies were resolved through discussion.
Data extraction
Data were extracted independently by one investigator and checked by a second independent investigator. A standard form was used to extract the following information: language of publication, country where study was conducted, year of publication, study design and objective(s), resources being compared, topic being searched, years the search covered, search strategies used, results (yield, recall, precision, reasons studies were missed), and author's conclusions. When data were not available, authors were contacted and asked to supply the missing information.
Data analysis
Data were analyzed using Splus 6.2 (Insightful Corporation 2003). Recall and precision were expressed as percentages with 95% confidence intervals which were calculated using exact methods [11]. Recall is the percentage of relevant studies that were identified by the search. Precision is the percentage of studies identified by the search that were relevant. Results were grouped by comparison (e.g., MEDLINE versus handsearching, MEDLINE vs. other reference standard). Meta-analysis was not performed due to large heterogeneity. Comparisons, however, were summarized using medians and ranges. With regards to the independence of the results, we conducted a sensitivity analysis around the inclusion of duplicate topics. Duplicate topics for exclusion were randomly selected and the median and range summaries were re-calculated.
Possible sources of heterogeneity were explored with numerical summaries both by within-study and between-study analyses. Within-study analyses are direct analyses and they occur when two searches are conducted with the same known conditions (i.e., strategy, expertise of the searcher, topic of the search, type of design) and some unknown conditions except for the condition or variable of interest. These are also called direct analyses. Between-study analyses are indirect analyses and are of a lower grade [12] in that there is a stronger potential for known variables (i.e., topic, strategy, author of search, topic of search, type of design) to confound results. The variables of interest we explored were: search strategy (Cochrane, Simple, Complex, Index), expertise of the searcher (Cochrane, librarian, non-librarian), and study design (RCT and CCT).
Searches were divided into the following four categories (modified from Hopewell et al. [8]): Complex (using a combination of types of search terms); Cochrane (the Cochrane Highly Sensitive Search Strategy (HSSS)); Simple (using five or fewer search terms which may include a combination of MeSH, Publication Type, keywords); and, Index (using one or two search terms (usually author's name or article title) to check/verify if the study is in a database).
Results
Description of included studies
Sixty-four studies met the criteria for inclusion in this analysis (Figure 1; see Additional file 1). Of these, 49 were published as full manuscripts, 12 were abstracts, 2 were letters, and 1 was a conference presentation. All studies were published between 1985 and 2003 with 94% being published after 1988, the same year MEDLINE became freely available through the PubMed interface . Approximately half the studies (n = 30) were conducted in the United Kingdom. Three studies were non-English (German, Dutch and Spanish). Thirty studies received funding, some from more than one source. Financial support was received from: 16 government programs, 2 pharmaceutical companies and 35 other sources (e.g., Universities, health trusts, foundations/associations, the National Library of Medicine and individual journals).
Figure 1 Quorum flow diagram.
The included studies searched a variety of topics which fell into four major categories: journal (e.g., Lancet, BMJ), disease/condition/state (e.g., hepatitis, rheumatoid arthritis), specialty/sub-specialty (e.g., rehabilitation, pediatrics) and methodology (e.g., search strategies) [see Additional file 1]. Generally, the objectives of the studies were: to compare different searches (e.g., handsearch vs. database); to determine recall/precision of a search; to handsearch a journal and check to see if trials were in a database; or to develop a trial register.
The reference standards varied (e.g., handsearching, handsearching plus MEDLINE, MEDLINE plus EMBASE plus other databases). The specific study design for which authors were searching varied by study: RCTs only (n = 27); RCTs and CCTs (n = 28); and RCTs, CCTs, and other designs (n = 9).
There were four major comparisons: MEDLINE vs. handsearch (n = 22), MEDLINE vs. MEDLINE + handsearch (n = 12), MEDLINE vs. other reference standard (n = 18), and EMBASE vs. reference standard (n = 13). There were 13 other comparisons with only one to two studies each (Table 1).
Table 1 Results
Comparison Recall Precision
Number of Studies† Number of Comparisons Median (Range) Number of Studies† Number of Comparisons Median (Range)
MEDLINE vs. Handsearching 23 41 22* 53 (7,97) 58 (7,97) 12 23 11 35 (0.03,99) 31 (0.03,78)
MEDLINE vs. MEDLINE + Handsearching 13 16 12* 70 (18,90) 75 (18,90) 6 8 5 49 (13,83) 40 (13,83)
MEDLINE vs. Reference Standard 13 24 18* 59 (17,98) 58 (17,98) 8 17 11 27 (1,91) 30 (1,91)
EMBASE vs. Handsearching 2 2 (42,88) 2 2 (9,17)
EMBASE vs. Reference Standard 9 14 13* 72 (13,100) 68 (13,100) 5 8 7 28 (0,48) 26 (0,48)
PsycINFO vs. Handsearching 2 2 (68,70) 2 2 (8,9)
PsycINFO vs. Reference Standard 2 4 2* 50 (0,65) (38,61) 2 3 2 39 (36,47) (36,47)
Trial Registries vs. Reference Standard 2 4 89 (84,95) 1 2 (96,97)
CINAHL vs. Reference Standard 2 2 (0,1) 1 1 0
Biosis vs. Reference Standard 1 1 47 1 1 1
CancerLIT vs. Handsearching 1 1 92 0 -- --
Cabnar vs. Reference Standard 1 1 47 1 1 2
CENTRAL vs. Reference Standard 1 1 78 0 -- --
Chirolars vs. Reference Standard 1 1 49 0 -- --
HealthStar vs. Reference Standard 1 1 53 1 1 1
Internet vs. Reference Standard 1 1 24 1 1 17
SciCitIndex vs. Reference Standard 1 1 61 0 -- --
* excludes duplicate topics
† excludes Index searches
Methodological quality of included studies
All studies indicated the type of study design for which they were searching and all but one specified the topic area. Most (70%) stated their inclusion/exclusion criteria. Eighty percent of the studies described (or indicated they were available) reproducible search strategies/methods. Half of the studies stated who developed the search strategies and of these only 2 did not provide reproducible information about their search strategy. Eighty-five percent of articles fully described how the reference standard was compiled. Twenty-five percent reported that at least 2 people independently screened searches and in 35% at least 2 people also independently applied eligibility criteria.
Quantitative results
Table 1 summarizes the results of the comparisons (e.g., MEDLINE versus handsearching, MEDLINE vs. other reference standard). Thirteen databases (including the Internet) were included in the numerical results. The results from 7 studies could not be used in the data analysis for the following reasons: did not use same journals for handsearch and MEDLINE [13]; insufficient data available (usually because it was an abstract) [14-18]; reporting flaw(s) which could not be clarified by author(s) [19].
MEDLINE
Forty-nine studies had usable data for the MEDLINE comparisons. Three comparisons were analyzed: MEDLINE versus handsearching (41 comparisons, 23 studies), MEDLINE versus MEDLINE plus handsearching (16, 13), and MEDLINE versus other reference standards (24, 13). Estimates of both recall and precision for all three comparisons varied substantially, ranging from 7 to 98% and 0.03 to 99%, respectively. The estimates for MEDLINE versus handsearching and MEDLINE versus a reference standard were comparable: median of 53 versus 59% for recall and 35 versus 27% for precision. Median recall and precision for MEDLINE versus MEDLINE plus handsearching were somewhat larger (70 and 49%, respectively).
EMBASE
Eleven studies had usable data for the EMBASE comparisons. Two comparisons were analyzed: EMBASE versus handsearching (2 comparisons, 2 studies) and EMBASE versus a reference standard (14, 9). Individual study estimates ranged from 13 to 100% for recall and 0 to 48% for precision. Summarizing all studies, medians were 65 and 72% for recall and, 13 and 28% for precision.
PsycINFO
Four studies contained data for the PsycINFO comparisons. Two comparisons were analyzed: PsycINFO versus handsearching (2 comparisons, 2 studies), and PsycINFO versus a reference standard (4, 2). Recall ranged from 0 to 70%. Precision ranged from 8 to 47%. Medians were 69 and 50% for recall and 9 and 39% for precision.
Other databases
Two studies investigated CINAHL and trial registers. Only one study had usable data from other databases (i.e., BIOSIS, CancerLit, CABNAR, CENTRAL, Chirolars, HealthSTAR, the Internet, SciCitIndex). The results from the trial registries versus a reference standard were consistent and high: 89% for median recall and 97% for median precision. The remaining comparisons ranged from 0 to 92% for median recall and 0 to 17% for median precision. Regardless of topic, there were too few included studies in these comparisons for this data to be representative.
Subgroup analyses
Table 2 shows the results for the direct subgroup analyses. Seven studies were included in the search strategy analysis. There were six comparisons of Simple versus Complex search strategies. All but one study [20] showed greater recall for the Complex search strategies. The trade-off which so often occurs between recall and precision did not occur: three out of the four Complex search strategies had larger (better) precision (not including Fergusson [20]).
Table 2 Direct Subgroup Analyses
Study Recall (95% CI) Precision (95% CI)
Simple versus Complex
Simple Complex Simple Complex
Adams[26] 1994 18 (15,21) 52 (48,56) 40 (35,46) 59 (55,63)
Bender[27] 1997 53 (47,58) 65 (59,70) -- 78 (73,83)
Dickerson[28] 1985a 18 (10,26) 29 (20,39) 65 (45,86) 72 (56,87)
Dickerson[28] 1985b 32 (15,50) 56 (38,74) 38 (19,57) 53 (35,70)
Fergusson[20] 2000 89 (81,97) 88 (80,96) 1 (1,1) 3 (2,4)
Marson[29] 1996 64 (55,73) 86 (80,93) 72 (63,81) 35 (29,40)
Simple versus Cochrane
Simple Cochrane Simple Cochrane
Brand[30] 1998a 59 93 -- --
Brand[30] 1998b 88 97 -- --
Fergusson[20] 2000 89 (81,97) 89 (81,97) 1 (1,1) 7 (6,9)
McDonald[31] unpub a 62 (55,68) 76 (70,82) -- --
McDonald[31] unpub b 52 (43,61) 91 (86,96) -- --
Complex versus Cochrane
Complex Cochrane Complex Cochrane
Fergusson[20] 2000 88 (80,96) 89 (81,97) 3 (2,4) 7 (6,9)
Author versus Librarian
Author Librarian Author Librarian
Kirpalani[21] 1989 34 (20,48) 53 (38,67) -- --
There were five direct comparisons of a Simple search strategy versus the Cochrane HSSS. Again, all but one of the comparisons [20] had larger sensitivities for the Cochrane search strategy. None of these four comparisons reported precision. Fergusson [20] found negligible differences for both recall and precision.
Three MEDLINE comparisons were considered for the indirect comparisons (Tables 3 and 4) since they were the only ones sizable in number. Our indirect results differed from the direct results. No systematic differences between the Simple and Complex search strategies in median recall were found: 49 versus 51% for MEDLINE versus handsearching, 48 versus 67% for MEDLINE versus MEDLINE plus handsearching, and 58 versus 40% for MEDLINE versus a reference standard. The precision results were similar: 76 versus 38%, 62 versus 51%, and 23 versus 35%, respectively. And although the median precision estimates for the Cochrane search were much smaller (9, 48, and 7%), the median recall estimates (67, 81, and 78%) were systematically greater when compared to the Simple and Complex search strategies.
Table 3 Indirect Subgroup Analyses: Recall
Subgroups MEDLINE vs. Handsearching MEDLINE vs. MEDLINE + Handsearching MEDLINE vs. Reference Standard
N Median (Range) N Median (Range) N Median (Range)
Search Strategy Index 6 93 (41,100) 0 1 66
Cochrane 12 67 (26,97) 7 81 (28,90) 6 78 (53,98)
Complex 15 51 (9,86) 6 67 (52,88) 10 40 (17,97)
Simple 14 49 (7,88) 3 48 (18,72) 8 58 (18,89)
Author Index 6 93 (41,100) 0 -- 1 66
Cochrane 12 67 (26,97) 7 81 (28,90) 6 78 (53,98)
Librarian 5 49 (20,62) 0 -- 3 56 (29,89)
Non-librarian 24 52 (7,88) 9 66 (18,88) 15 48 (17,97)
Design RCT 19 54 (9,97) 1 28 10 79 (25,97)
CCT 28 62 (7,100) 15 72 (18,90) 15 56 (17,98)
Table 4 Indirect Subgroup Analyses: Precision
Subgroups MEDLINE vs. Handsearching MEDLINE vs. MEDLINE + Handsearching MEDLINE vs. Reference Standard
N Median (Range) N Median (Range) N Median (Range)
Search Strategy Cochrane 7 9 (0.03,76) 3 48 (13,49) 3 7 (1,7)
Complex 12 38 (22,78) 3 51 (13,59) 9 35 (3,91)
Simple 4 76 (68,99) 2 (40,83) 5 23 (1,65)
Author Cochrane 7 9 (0.03,76) 3 48 (13,49) 3 7 (1,7)
Librarian 1 34 0 -- 2 (53,72)
Non-librarian 15 50 (22,99) 5 51 (13,83) 12 29 (1,91)
Design RCT 12 43 (22,79) 0 -- 7 7 (1,91)
CCT 11 28 (0.03,99) 8 49 (13,83) 10 33 (7,72)
Only one study [21] directly compared search strategies from two different authors (i.e., a librarian versus a non-librarian). In this one example, the librarian's Complex search had a recall of 53% and the non-librarian's Complex search had a recall of 34. For the indirect subgroup comparisons, the librarians did not systematically outperform the non-librarians on either median recall or median precision; however the Cochrane HSSS (as author) did outperform the librarians and non-librarians on median recall (67, 81 and 78%).
No studies directly compared searching for RCTs versus CCTs. Based on indirect comparisons, the three MEDLINE comparisons showed no systematic difference in median recall and precision between design types.
A sensitivity analysis excluding duplicate topics was performed due to the concern for non-independence between studies. Studies or comparisons searching on the same topic may include some of the same relevant studies. We picked one comparison randomly from each topic area and eliminated it from the main quantitative results. The results are shown in Table 1. We found that recall had similar ranges and medians signifying that non-independence was not distorting our results.
Reasons for missed trials
Table 5 lists the most common reasons articles were missed in both the electronic and handsearches. Forty-two studies reported reasons for missing studies from the search or handsearch. For electronic databases, the reason cited most often (67%) for missed studies was inadequate or inappropriate indexing. Other major reasons why studies were not found in databases included: they were published as abstracts, books, book reviews, brief reports, letters, proceedings or supplements, etc. (i.e., grey literature) (21%); keywords or methodology were not reported by author (21%); insufficient or restricted search strategy (14%); article(s) were omitted or missing from a resource (14%).
Table 5 Reasons why studies were missed by electronic search and handsearch
Category Reason Studies citing reason (n = 42)
Electronic Resources
Indexing
Inadequate indexing in general 16[3,27,29,32–44]
Journal not indexed in resource 8[3,13,20,27,35,40,45,46]
Not indexed with proper methodologic terms 9[21,26,28,29,32,36,42,47,48]
Not indexed as RCT/CCT 7[35,37,49–53]
Not indexed for any of the methodologic terms used by searchers 2[21,54]
Database does not contain relevant index terms 2[17,50]
Not indexed with any of the subheadings used in searches 1[21]
Search Strategy
Insufficient or restricted search strategy 6[27,32,35,55–57]
Reporting
Keywords or methodology not reported by author 9[13,19,29,34,41,48,53,58,59]
Database
Issue(s) omitted from resource 4[3,26,56,60]
Time lag in updating resource 2[60,61]
Article(s) omitted or missing from resource 6[3,29,36,50,51,56]
Journal
Published before database was created or beyond coverage years of database 2[56,60]
Other
Unknown reason 6[26,38,38,54,60,62]
Published as abstracts, books, book reviews, brief reports, letters, proceedings, supplements, etc. 9[20,26,38,38,49,51,56,63]
• Wrong abstract assigned to reference[36]
• Journals do not encourage authors to explicitly report methodology[37]
• Length and complexity of search is limited using search engines[64]
• Ongoing or recently finished studies[64]
• Found through references[20]
• Missed most Japanese reports[65]
• Some personal communication did not get responses; organizations and people must be willing to supply trial information[18]
• Not all abstracts are published in full[50]
• Authors misspelled gingko[39]
• Many RCTs are not identifiable in database[66]
Handsearch
Handsearchers not trained properly 2[57,67]
Methodologic terms "hidden" in article 1[26]
Searcher fatigue, boredom 1[26]
Journal not handsearched 2[35,42]
Article not an RCT or CCT 1[54]
Other
Different topic than what handsearchers were looking for 1[42]
In this study, seven studies performed Index searches. Six comparisons using the Index searches were found for MEDLINE versus handsearching. The median recall was 93% and the range was 41 to 100%. On average, 7% of studies were not indexed adequately. One study compared MEDLINE to a reference standard, their Index search produced 66% of the included studies. Two further Index searches were performed: EMBASE versus handsearching and PsycINFO versus handsearching; their recalls were 52 and 97%, respectively.
For handsearching, few authors provided information for why trials were missed. Handsearches had high precision and some studies did not miss any references through their handsearches. In the 3 studies where handsearchers missed studies, authors reported the reasons for missing studies were the handsearchers were not trained properly or they had fatigue/boredom. In two studies where trials were missed, authors reported that the journal was not handsearched, yet a database was used to search for this same journal. One of the missed articles was misclassified by a handsearcher as an RCT/CCT and one had a different topic than what handsearchers were meant to identify. Results from the MEDLINE versus MEDLINE plus handsearching comparisons quantify the percentage of trials missed by handsearching. In 13 studies, the median percentage found in MEDLINE but not by handsearching was 6% (range 1 to 15%).
Discussion
For certain topics trial registries may be sufficient (e.g., perinatology, Japanese), however, the median recall estimates (Table 1) were not large enough to support single-source searches. These data highlight the importance of searching multiple sources when conducting a systematic review. Initiatives to compile references from different sources, such as CENTRAL and other trial registries, need to receive continued encouragement and support in order to eliminate the need for multiple-source search endeavors.
Over and above the recalls, the median precisions are quite low and indicate a need for improved indexing in databases. Efforts to improve and standardize the indexing of various databases need to be supported. Guidelines for journals and authors regarding the reporting of key methodological or subject terms when publishing studies would facilitate these efforts [22]. In addition to very poor precisions, the authors of our included studies reported precisions or the data necessary to calculate precision only 40% (47/117 comparisons) of the time.
Most of the research has involved MEDLINE and EMBASE, two of the major databases that the systematic review community recommends reviewers search. However, searching multiple databases can be difficult, time consuming and costly. For example, although MEDLINE is available freely on the Internet through PubMed, EMBASE is very costly and many institutions do not subscribe to it. This is of particular concern as studies have demonstrated that there is 17 to 75% overlap between MEDLINE and EMBASE [3,23,24] indicating that EMBASE may yield a substantial number of unique articles. To date, the gold standard for conducting systematic reviews still remains searching multiple bibliographic databases and hand searching. In addition, using only MEDLINE for systematic reviews still results in important trials being missed thereby compromising the external validity of the review.
Optimally, it would be most efficient to search few resources, retrieve a maximum yield of relevant trials, and retrieve a minimum yield of irrelevant trials. The Cochrane Collaboration is trying to achieve this with the Controlled Trials Register (CENTRAL) available through the Cochrane Library. The register now includes over 420,000 RCTs and CCTs. While there are numerous studies that discuss the vast amount of trials that have been added to CENTRAL through handsearching efforts, there are very few studies evaluating whether CENTRAL can be searched exclusively for RCT/CCTs. If one resource (e.g., CENTRAL) can be searched to identify RCT/CCTs, this would substantially reduce the time and costs associated with searching.
There was extensive heterogeneity among topics investigated in the studies included in this review. For the comparisons which had many studies, the values for both recall and precisions covered most of the possible range (e.g., 0–100). Thus, the topic searched may be the strongest determinant of the results. Topics are indexed differentially within and across various sources. Due to the between-study heterogeneity, very little can be concluded about the indirect subgroup results.
Over the 20 years that this review covers, it was noted that the older studies were conducted prior to indexing improvements in resources, especially MEDLINE. While there have been numerous changes in search technology in the past two decades, upon conducting post-hoc sub-group analyses, no difference was found. In addition, a sub-group analysis was done of recent studies and no difference was found when compared to the results of older studies. As mentioned above, this may be due to topic heterogeneity, not changing search technology. Thus, including the older studies did not confound our results and did not lead to the conclusion that there is one sufficient resource which identifies RCT/CCTs. Unfortunately, the more recent studies are not showing results which significantly differ from the ones obtained 20 years ago.
We found that, generally, both Complex and Cochrane search strategies performed better in recall than did Simple search strategies without loss in precision. However, the indirect subgroup results for recall showed support for this finding for the Cochrane search strategies, but not for the Complex search strategies. The Cochrane search strategy precisions were poorer in the indirect subgroup results, however data were sparse. Little direct evidence was available comparing searchers with different expertise. No direct evidence was available comparing searches for different design types. Without supporting direct subgroup evidence, conclusions from the indirect subgroup evidence would be too speculative. Other reasons that may explain the heterogeneity include: 1) the time period covered by the search (indexing as well as other search technologies have progressed over time which would affect the accuracy of searches); and 2) the methodological quality of the study. For example the rigor with which handsearching was done, searches were screened, or inclusion criteria applied (e.g., having 2 people independently perform each step) would affect the comparability of results across studies.
The quality of the included studies varied and, in most cases, the poor quality result was due to the lack of rigor in the reporting of the selection methodology. One-third reported that standard inclusion and exclusion criteria were developed and applied to each database/method at the relevance stage. Almost all studies did not indicate that at least 2 people independently screened the searches for potentially relevant studies. In addition, two-thirds of the included studies did not indicate that at least 2 people independently applied eligibility criteria to identify relevant studies. Quality of these studies can be improved by adopting more stringent methodology and reporting its use.
Post-hoc, we looked at how the calculated results of recall and precision may have improved over the last two decades considering the changes in search technology (in particular, indexing). Within our three MEDLINE comparisons, we found no pattern of association between year of publication and results. The correlations ranged from -0.91 to 1. As mentioned above, this quantitative analysis may be too dilute due to topic heterogeneity. We suggest a within topic analysis to robustly test for improvements in search technology.
This paper provides the most current and comprehensive review of the existing evidence comparing any electronic database against any other source or combination of sources. This and previous reviews demonstrate that there is a dearth of evidence regarding the use of different databases to retrieve RCTs, with the notable exception of handsearching and MEDLINE. Therefore, searching multiple resources to retrieve RCTs cannot be ruled out based upon this evidence. There needs to be more research done on major databases such as: EMBASE, CENTRAL, PsycINFO and trial registries in order to gather more information about the value of these databases in identifying RCT/CCTs for various topical areas. This review is more comprehensive than previous work in this area [8] and reflects the different ways that searches can be conducted (i.e., using a variety of databases and types of searches). Moreover, while a previous review focused upon between-study subgroup comparisons [8] (e.g., Complex versus Simple search strategies), we also systematically examined the within-study subgroup comparisons which provides more valid information [12]. However, similar to negative clinical trials, it is important recognize the limitations of current resources and the implications for decision-making.
Limitations
There are several limitations to this study. Foremost, there is a lack of a validated quality score for this type of study (i.e., comparative). Reference standards are difficult to compare as they are generally different and may not be reported in enough detail to be reproducible. As well, the topic chosen to search can determine the success of the strategy. In addition, there are limitations to using precision and recall which are addressed by Kagolovsky and Moehr [25].
Conclusion
Implications for practice
Since recall is low with single resources, multiple-source comprehensive searches continue to be necessary. The Cochrane search strategy or Complex search strategy in consultation with a librarian are recommended.
Implications for research
Efforts to enhance and build CENTRAL, a large trial registry, need to be continued. A number of the resources used to find trials for CENTRAL (e.g., journals, grey literature) are not indexed in MEDLINE, therefore CENTRAL has a significant amount of unique information not found in any other source. CENTRAL is also free for researchers in developing countries and available in CD-ROM and on the internet. Based upon the results of why studies were missed, indexing efforts also need to improve. Guidelines should be provided for authors to include MeSH terms and keywords in their abstracts which can then be used by indexers. Other resources that need to be studied include: Trial registries, LILACS, PsycINFO, Science Citation Index, BIOSIS, CABNAR and CINAHL. In addition, those researchers studying searches need to report precision results.
List of abbreviations used
CCT – controlled clinical trial
RCT – randomized controlled trial
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
ETC conceived of the study, designed and coordinated it and drafted the manuscript. NW performed the statistical analysis and drafted and read the final manuscript. LH and KC participated in the study and helped to draft the manuscript. TK participated in the study design and read the final manuscript. All authors participated in the design of the study and read and approved the final manuscript.
Appendix 1: MEDLINE Search Strategy
1 medline.mp.
2 internet.mp.
3 embase.mp.
4 (psyclit or psycinfo or psychlit or psychinfo).mp.
5 "web of science".mp.
6 cinahl.mp.
7 sigle.mp.
8 "system for information on grey literature in europe".mp.
9 lilacs.mp.
10 excerpta medica.mp.
11 "science citation index".mp.
12 "science citation abstracts".mp.
13 scisearch.mp.
14 toxline.mp.
15 aidsline.mp.
16 cancerline.mp.
17 pubmed.mp.
18 grateful med.mp.
19 cabnar.mp.
20 "health star".mp.
21 healthstar.mp.
22 "current contents".mp.
23 "cochrane library".mp.
24 ("cochrane controlled trials register" or central or cctr).mp.
25 "database of abstracts of reviews of effectiveness".mp.
26 eric.tw.
27 "world wide web".mp.
28 dissertation$.mp.
29 thesis.mp.
30 "institute of scientific information".mp.
31 isi.mp.
32 "inside information plus".mp.
33 firstsearch.mp.
34 "international pharmaceutical abstracts".mp.
35 "biological abstracts".mp.
36 (dare and cochrane).mp.
37 pascal.tw.
38 or/1–36
39 search$.mp.
40 (handsearch$ or "hand search$").mp.
41 compar$.mp.
42 "manual search$".mp.
43 or/39–42
44 (controlled adj2 trial$).mp.
45 clinical trial$.mp.
46 (randomized controlled trial$ or randomised controlled trial$).mp.
47 (rct or cct).mp.
48 or/44–47
49 and/38,43,48
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
Characteristics of included studies, information about the studies used in this systematic review
Click here for file
Acknowledgements
We are indebted to ARCHE who provided financial assistance for this project when ETC and NW formerly worked with them. Thank you to Research Assistants Michelle Tubman and Philip Berry for assisting with data collection and study retrieval.
==== Refs
Sampson M Barrowman NJ Moher D Klassen TP Pham B Platt R St John PD Viola R Raina P Should meta-analysis search EMBASE in addition to MEDLINE? J Clin Epidemiol 2003 56 943 955 14568625 10.1016/S0895-4356(03)00110-0
Royle P Waugh N Literature searching for clinical and cost-effectiveness studies used in health technology assessment reports (TAR) carried out for the National Institute for Clinical Excellence appraisal system Health Technol Assess 2003 7 1 64
Suarez-Almazor ME Belseck E Homik J Dorgan M Ramos-Remus C Identifying clinical trials in the medical literature with electronic databases: MEDLINE alone is not enough Control Clin Trials 2000 21 476 487 11018564 10.1016/S0197-2456(00)00067-2
Alderson P, Green S, Higgins JPT Locating and selecting studies for reviews In: The Cochrane Library, Issue 1, 2004 Cochrane Reviewers' Handbook 422 [updated December 2003] 2004 Chichester, UK: John Wiley & Sons, Ltd
Allen IE Olkin I Estimating time to conduct a meta-analysis from number of citations retrieved JAMA 1999 282 634 635 10517715 10.1001/jama.282.7.634
Alderson P, Green S, Higgins JPT Locating and selecting studies for reviews In: The Cochrane Library, Issue 1, 2004 Cochrane Reviewers' Handbook 422 [updated December 2003] 2004 Chichester, UK: John Wiley & Sons, Ltd
Dickersin K Manheimer E Wieland S Robinson KA Lefebvre C McDonald S Development of the Cochrane Collaboration's CENTRAL Register of controlled clinical trials Eval Health Prof 2002 25 38 64 11868444 10.1177/0163278702025001004
Hopewell S Clarke M Lefebvre C Scherer R Handsearching versus electronic searching to identify reports of randomized trials (Cochrane Methodology Review) The Cochrane Library 2004 Chichester, UK: John Wiley & Sons, Ltd
Jadad AR Moore RA Carroll D Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996 17 1 12 8721797 10.1016/0197-2456(95)00134-4
Lijmer JC Mol BW Heisterkamp S Bonsel GJ Prins MH van der Meulen JH Bossuyt PM Empirical evidence of design-related bias in studies of diagnostic tests JAMA 1999 282 1061 1066 10493205 10.1001/jama.282.11.1061
Fleiss JL Statistical Methods for Rates and Proportions 1981 New York, NY: John Wiley & Sons Ltd
Oxman AD Guyatt GH Validation of an index of the quality of review articles J Clin Epidemiol 1991 44 1271 1278 1834807 10.1016/0895-4356(91)90160-B
Silagy C Developing a register of randomised controlled trials in primary care BMJ 1993 306 897 900 8490419
Tsutani K Sakuma A How to access RCTs in Japan through network access from abroad. [poster] Proceedings of the 2nd Annual Cochrane Colloquium October 1994; Canada
Thomas B Gubitz G McInnes A Krabshuis J Counsell C A multifile search of five bibliographic databases for stroke trials. [poster] Proceedings of the 6th Annual Cochrane Colloquium October 1998; USA
Hedger NA Croft AMJ Rowe M Handsearching the Journal of the Royal Naval Medical Service for trials J R Nav Med Serv 1999 85 108 111 10707452
Veldhuyzen van Zanten SJO Cleary C Talley NJ Peterson TC Nyren O Bradley LA Verlinden M Tytgat GNJ Drug treatment of functional dyspepsia: a systematic analysis of trial methodology with recommendations for design of future trials Am J Gastroenterol 1996 91 660 673 8677926
Helmer D Savoie I Green C Kazanjian A Evidence-based practice: extending the search to find material for the systematic review Bull Med Libr Assoc 2001 89 346 352 11837256
Kennedy G Rutherford G Identifying randomized controlled trials in the journal Aids Proceedings of the 8th Annual Cochrane Colloquium October 2000; South Africa
Fergusson D Laupacis A Salmi LR McAlister FA Huet C What should be included in meta-analyses? An exploration methodological issues using the ISPOT meta-analyses International Journal of Technology Assessment in Health Care 2000 16 1109 1119 11155831 10.1017/S0266462300103150
Kirpalani H Schmidt B McKibbon KA Haynes RB Sinclair JC Searching MEDLINE for randomized clinical trials involving care of the newborn Pediatrics 1989 83 543 546 2648304
Stroup DF Berlin JA Morton SC Olkin I Williamson GD Rennie D Moher D Becker BJ Sipe TA Thacker SB Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group JAMA 2000 283 2008 2012 10789670 10.1001/jama.283.15.2008
Minozzi S Pistotti V Forni M Searching for rehabilitation articles on MEDLINE and EMBASE. An example with cross-over design Arch Phys Med Rehabil 2000 81 720 722 10857512
Topfer LA Parada A Menon D Noorani H Perras C Serra-Prat M Comparison of literature searches on quality and costs for health technology assessment using the MEDLINE and EMBASE databases Int J Technol Assess Health Care 1999 15 297 303 10507189
Kagolovsky Y Moehr JR Current status of the evaluation of information retrieval J Med Syst 2003 27 409 424 14584618 10.1023/A:1025603704680
Adams CE Power A Frederick K Lefebvre C An investigation of the adequacy of MEDLINE searches for randomized controlled trials (RCTs) of the effects of mental health care Psychol Med 1994 24 741 748 7991756
Bender JS Halpern SH Thangaroopan M Jadad AR Ohlsson A Quality and retrieval of obstetrical anaesthesia randomized controlled trials Can J Anaesth 1997 44 14 18 8988818
Dickersin K Hewitt P Mutch L Chalmers I Chalmers TC Perusing the Literature: Comparison of MEDLINE searching with a perinatal trials database Control Clin Trials 1985 6 306 317 3907973 10.1016/0197-2456(85)90106-0
Marson AG Chadwick DW How easy are randomized controlled trials in epilepsy to find on MEDLINE? The sensitivity and precision of two MEDLINE searches Epilepsia 1996 37 377 380 8603644 10.1111/j.1528-1157.1996.tb00575.x
Brand M Gonzalez J Aguilar C Identifying RCTs in MEDLINE by publication type and through the Cochrane Strategy: the case in hypertension 1988 poster B22
McDonald SJ Lefebvre C Clarke MJ Identifying reports of controlled trials in the BMJ and The Lancet BMJ 1996 313 1116 1117 8916697
Bernstein F The retrieval of randomized clinical trials in liver diseases from the medical literature: manual versus MEDLARS searches Control Clin Trials 1988 9 23 31 3281794 10.1016/0197-2456(88)90006-2
Clarke M McDonald S Lefebvre C Identifying reports of randomized controlled trials (RCTs) in the Lancet: the contribution of the Cochrane Collaboration. [poster] Proceedings of the 4th Annual Cochrane Colloquium October 1996; Australia
Cullum N Identification and analysis of randomised controlled trials in nursing: a preliminary study Qual Health Care 1997 6 2 6 10166598
Dickersin K Scherer R Lefebvre C Identifying relevant studies for systematic reviews BMJ 1994 309 1286 1291 7718048
Duggan LM Prevalence study of the randomized controlled trials in the Journal of Intellectual Disability Research: 1957–1994 J Intellect Disabil Res 1997 41 232 237 9219072 10.1046/j.1365-2788.1997.02626.x
Dumbrigue HB Esquivel JF Jones JS Assessment of MEDLINE search strategies for randomized controlled trials in prosthodontics J Prosthodont 2000 9 8 13 11074022 10.1111/j.1532-849X.2000.00008.x
Jadad AR Carroll D Moore A McQuay H Developing a database of published reports of randomised clinical trials in pain research Pain 1996 66 239 246 8880846 10.1016/0304-3959(96)03033-3
Kleijnen J Knipschild P The comprehensiveness of MEDLINE and EMBASE computer searches. Searches for controlled trials of homoeopathy, ascorbic acid for common cold and ginkgo biloba for cerebral insufficiency and intermittent claudication Pharm Weekbl Sci 1992 14 316 320 1437515
Marti J Bonfill X Urrutia G Lacalle JR Bravo R [The identification and description of clinical trials published in Spanish journals of general and internal medicine during the period of 1971–1995]. [Spanish] Medicina Clinica 1999 112 28 34 10618797
Nwosu CR Khan KS Chien PW A two-term medline search strategy for identifying randomized trials in obstetrics and gynecology Obstet Gynecol 1998 91 618 622 9540953 10.1016/S0029-7844(97)00703-5
Poynard T Conn HO The retrieval of randomized clinical trials in liver disease from the medical literature: a comparison of MEDLARS and manual methods Control Clin Trials 1985 6 271 279 3907971 10.1016/0197-2456(85)90103-5
Schlomer G RCTs and systematic reviews in nursing literature: a comparison of German and international nursing research [German] Pflege, 12(4):250-8, 1999 Aug 1999 12 250 258
Watson RJ Richardson PH Identifying randomized controlled trials of cognitive therapy for depression: comparing the efficiency of EMBASE, MEDLINE and PsycINFO bibliographic databases Br J Med Psychol 1999 72 535 542 10616135 10.1348/000711299160220
Bassler D Galandi D Forster J Antes G Handsearching in German paediatric journals.[poster] Proceedings of the 8th Annual Cochrane Colloquium October 2000; South Africa
Bereczki D Gesztelyi G A Hungarian example for handsearching specialized national healthcare journals of small countries for controlled trials. Is it worth the trouble? Health Libr Rev 2000 17 144 147 11186806 10.1046/j.1365-2532.2000.00280.x
Galandi D Bassler D Antes G Identifying randomized controlled trials published in German general health care journals using Medline and Embase: How useful is the controlled vocabulary? [poster] Proceedings of the 8th Annual Cochrane Colloquium October 2000; South Africa
Jadad AR McQuay HJ A high-yield strategy to identify randomized controlled trials for systematic reviews Online J Curr Clin Trials 1993 Doc No 33 3973
Adetugbo K Williams H How well are randomized controlled trials reported in the dermatology literature? Arch Dermatol 2000 136 381 384 10724201 10.1001/archderm.136.3.381
Hopewell S Clarke M Lusher A Lefebvre C Westby M A comparison of handsearching versus MEDLINE searching to identify reports of randomized controlled trials Stat Med 2002 21 1625 1634 12111923 10.1002/sim.1191
Olesen V Engell L Jensen KL Gotzsche PC Randomised clinical trials in the Scandinavian Journal of Rheumatology Scand J Rheumatol 2000 29 349 351 11132202 10.1080/030097400447534
Pasternack I Varonen H Juva K Lehtinen E Jormanainen V Makela M Hand Searching of Finnish Medical Journals. [poster] Proceedings of the 5th Annual Cochrane Colloquium October 1997; Netherlands
Solomon MJ Laxamana A Devore L McLeod RS Randomized controlled trials in surgery Surgery 1994 115 707 712 8197563
Smith PJ Moffatt ME Gelskey SC Hudson S Kaita K Are community health interventions evaluated appropriately? A review of six journals J Clin Epidemiol 1997 50 137 146 9120506 10.1016/S0895-4356(96)00338-1
Brazier H Murphy AW Lynch C Bury G Wilson S Searching for the evidence in pre-hospital care: a review of randomised controlled trials J Accid Emerg Med 1999 16 18 23 9918280
Gotzsche PC Lange B Comparison of search strategies for recalling double-blind trials from MEDLINE Dan Med Bull 1991 38 476 478 1802637
Langham J Thompson E Rowan K Identification of randomized controlled trials from the emergency medicine literature: comparison of hand searching versus MEDLINE searching Ann Emerg Med 1999 34 25 34 10381991
Gluud C Nikolova D Quality assessment of reports on clinical trials in the Journal of Hepatology J Hepatol 1998 29 321 327 9722217 10.1016/S0168-8278(98)80021-4
Watson RJD Richardson PH Accessing the literature on outcome studies in group psychotherapy: the sensitivity and precision of Medline and PsycINFO bibliographic database searching Br J Med Psychol 1999 72 127 134 10194578 10.1348/000711299159763
Hay PJ Adams CE Lefebvre C The efficiency of searches for randomized controlled trials in the International Journal of Eating Disorders: a comparison of handsearching, EMBASE and PsycLIT Health Libr Rev 1996 13 91 96 10.1046/j.1365-2532.1996.1320091.x
Avenell A Handoll HHG Grant AM Lessons for search strategies from a systematic review, in the Cochrane Library, of nutritional supplementation trials in patients after hip fracture Am J Clin Nutr 2002 73 505 510 11237924
Ahmed I Souza Soares KV Seifas R Adams CE Randomized controlled trials in Archives of General Psychiatry (1959–1995): a prevalence study Arch Gen Psychiatry 1998 55 754 755 9707389 10.1001/archpsyc.55.8.754
Croft AM Vassallo DJ Rowe M Handsearching the Journal of the Royal Army Medical Corps for Trials J R Army Med Corps 1999 145 86 88 10420346
Eysenbach G Tuische J Diepgen TL Evaluation of the usefulness of Internet searches to identify unpublished clinical trials for systematic reviews Med Inform (Lond) 2001 26 203 218
Furukawa TA Inada T Adams CE McGuire H Inagaki A Nozaki S Are the Cochrane group registers comprehensive? A case study of Japanese psychiatry trials BMC Med Res Methodol 2002 2
Neal B Rodgers A Mackie MJ MacMahon S Forty years of randomised trials in the New Zealand Medical Journal N Z Med J 1996 109 372 373 8890875
Langham J Thompson E Rowan K Randomised controlled trials from the critical care literature: identification and assessment of quality Clin Intensive Care 2002 13 73 83
|
16092960
|
PMC1232852
|
CC BY
|
2021-01-04 16:32:52
|
no
|
BMC Med Res Methodol. 2005 Aug 10; 5:24
|
utf-8
|
BMC Med Res Methodol
| 2,005 |
10.1186/1471-2288-5-24
|
oa_comm
|
==== Front
BMC Musculoskelet DisordBMC Musculoskeletal Disorders1471-2474BioMed Central London 1471-2474-6-461613732610.1186/1471-2474-6-46Research ArticleQuadriceps force generation in patients with osteoarthritis of the knee and asymptomatic participants during patellar tendon reflex reactions: an exploratory cross-sectional study Dixon John [email protected] Tracey E [email protected] Teesside Centre for Rehabilitation Sciences, University of Teesside, The James Cook University Hospital, Middlesbrough, UK2 HealthQWest, Glasgow Caledonian University, Cowcaddens Road, Glasgow, UK2005 1 9 2005 6 46 46 2 3 2005 1 9 2005 Copyright © 2005 Dixon and Howe; licensee BioMed Central Ltd.2005Dixon and Howe; 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 postulated that muscle contraction is slower in patients with osteoarthritis of the knee than asymptomatic individuals, a factor that could theoretically impair joint protection mechanisms. This study investigated whether patients with osteoarthritis of the knee took longer than asymptomatic participants to generate force during reflex quadriceps muscle contraction. This was an exploratory study to inform sample size for future studies.
Methods
An exploratory observational cross sectional study was carried out. Two subject groups were tested, asymptomatic participants (n = 17), mean (SD) 56.7 (8.6) years, and patients with osteoarthritis of the knee, diagnosed by an orthopaedic surgeon, (n = 16), age 65.9 (7.8) years. Patellar tendon reflex responses were elicited from participants and measured with a load cell. Force latency, contraction time, and force of the reflex response were determined from digitally stored data. The Mann-Whitney U test was used for the between group comparisons in these variables. Bland and Altman within-subject standard deviation values were calculated to evaluate the measurement error or precision of force latency and contraction time.
Results
No significant differences were found between the groups for force latency (p = 0.47), contraction time (p = 0.91), or force (p = 0.72). The two standard deviation measurement error values for force latency were 27.9 ms for asymptomatic participants and 16.4 ms for OA knee patients. For contraction time, these values were 29.3 ms for asymptomatic participants and 28.1 ms for OA knee patients. Post hoc calculations revealed that the study was adequately powered (80%) to detect a difference between the groups of 30 ms in force latency. However it was inadequately powered (59%) to detect this same difference in contraction time, and 28 participants would be required in each group to reach 80% power.
Conclusion
Patients with osteoarthritis of the knee do not appear to have compromised temporal parameters or magnitude of force generation during patellar tendon reflex reactions when compared to a group of asymptomatic participants. However, these results suggest that larger studies are carried out to investigate this area further.
==== Body
Background
Osteoarthritis (OA) of the knee is associated with quadriceps muscle weakness [1,2], muscle dysfunction [3], and proprioceptive impairments [4] that may contribute to the pathogenesis or progression of OA knee by the production of increased joint damage. Minor neuromuscular incoordination has been termed "microklutziness" [5], and may result in impulsive joint loading and an increased heel strike force [1,5,6]. As the quadriceps muscle group is a main stabiliser of the knee joint, muscle weakness or atrophy will of course reduce the amount of protective force generated at the knee joint [1]. In addition, however, if the speed of muscle contraction is also affected and slower, then it will also take longer for protective and stabilising muscle contraction to occur [1,7-9]. Marks et al. [8] observed that the ability to generate force quickly during voluntary muscle contraction was impaired in the quadriceps of OA knee patients. However, due to the protective reflex mechanisms that operate around the knee joint [3,7,10], muscle force generation during reflex reactions may be at least as or more important than voluntary contractions [7,11]. There is an absence of research on quadriceps reflex force generation in OA knee, which may be vital in these protective reflexes. This knowledge may be useful in understanding the aetiology of OA knee. Furthermore, because exercise may improve the speed of force generation [12], and thus may improve knee joint protection [3,9], information on reflex force generation may allow rehabilitative and preventative measures to be improved for this population.
The aim of this study was to investigate whether reflex force generation was impaired in the quadriceps of OA knee patients compared to asymptomatic participants. This was achieved by measuring the standard temporal parameters termed force latency (FL) and contraction time (CT) [13,14], and force during the patellar tendon reflex. FL is the time from tendon tap to onset of quadriceps force generation, and CT is the time from force onset to peak force. Our experimental hypothesis was that there would be a difference in FL, CT and force between the groups. As no published data were available on the parameters of interest in OA knee patients, data from this preliminary study will inform sample size calculations for any future studies.
Methods
An exploratory observational cross sectional study was carried out in conjunction with an EMG investigation [15].
Subjects
Ethical approval was granted by the local research ethics committee. Our sample were opportunistic. All subjects gave written and verbal informed consent before taking part in the study. Two groups were tested, symptomatic OA knee patients and asymptomatic subjects. The descriptive characteristics of the subjects are shown in Table 1. OA patients were recruited from South Tees Hospitals NHS Trust, UK, outpatients orthopaedic clinics. Diagnosis of OA knee was made by an orthopaedic surgeon according to the American College of Rheumatology criteria, [16] using clinical signs and symptoms and the presence of osteophytes determined by weight-bearing radiographs. Asymptomatic subjects comprised a convenience sample of volunteers recruited from hospital and university sites and local clubs, and were individuals who reported having no history of knee pain.
Table 1 Descriptive characteristics of participants
Group Age (years) Mean (SD) Height (m) Mean (SD) Mass (kg) Mean (SD) BMI (kg/m2) Mean (SD) Sex (M:F)
Asymptomatic 56.7 (8.6) 1.68 (0.09) 74.4 (10.9) 26.3 (2.7) 8:9
OA 65.9 (7.8) 1.68 (0.10) 82.1 (11.2) 29.0 (2.4) 11:5
Subjects were excluded if they presented with significant cognitive, cardiorespiratory, neurological, or musculoskeletal impairments (excepting OA knee in the patient group) that limited functional ability, or reported use of medication affecting neuromuscular function. One OA knee patient was excluded from the study, as a detectable reflex response could not be elicited, reducing the number of patients from 17 to 16 in that group.
Materials and procedure
For testing, each subject was seated on an adjustable padded 'quadriceps chair' [17-19] with the posterior aspect of the knee positioned at the front edge of the seat and arms folded across the chest. In patients with unilateral OA, the symptomatic limb was tested. In bilateral OA knee patients and asymptomatic subjects, data were recorded from the dominant limb, which was defined as the limb with which they would kick a ball.
Measurements were made in a non-weight bearing open kinetic chain manner [20,21]. A comfortable non-extensible strap was positioned around the lower leg with the bottom edge at the level of the lateral malleolus. This strap was attached to a light chain that passed horizontally to a calibrated load cell type TC601 (Amber Instruments Ltd, UK) attached to the chair, but adjustable for height and direction. For each subject, the knee joint was carefully set at 90° of flexion, by adjustment of the non-extensible strap, with the chain passing horizontally without slack to the load cell. The load cell output was amplified using a load cell amplifier type LAU 64.200 (Sensor techniques Ltd, UK). The amplified output from this was input to an MP100 workstation (BIOPAC Inc., USA), recorded and stored digitally.
All subjects were blindfolded to minimize external influences. A standard tape of classical music was played to them through headphones. The patellar tendon reflex was elicited by a patellar hammer connected to a pivot on a specially constructed height-adjustable rig [17]. This ensured reproducibility of the force and strike location of each stimulus. The force produced by the hammer was repeatedly tested using the load cell, and was a constant 4.2 N [17]. It was felt that this force was sufficient to elicit a reflex response, and that higher force could cause pain to patients. A vibration sensor (Maplin Electronics, UK) attached to the hammer arm and connected to the MP100 workstation produced a square wave time marker on the recording on impact. Ten taps to the patellar tendon were delivered at random time intervals (10 to 20 s apart) to reduce habituation or fatigue [22,23]. Data were sampled at 2048 Hz using a physiological data acquisition system (BIOPAC Inc., USA) comprising the MP100 workstation with a high level transducer HLT100 and dedicated analysis software (AcqKnowledge 3.5.3).
Force latency, defined as the time from tendon tap to onset of force generation, and contraction time, defined as the time from force onset to peak force, were determined from each response using the digital outputs of the AcqKnowledge software. The peak force of each reflex response was also recorded. For each subject the resting tension level was taken into account by subtraction when calculating the force output. The mean values of the ten responses per subject for each variable were analysed using SPSS version 9.0. As statistical analysis showed that the data for FL and CT and force did not follow a normal distribution, (Shapiro-Wilk test p < 0.05), the between group comparisons were made with the non-parametric Mann-Whitney U test and non-parametric approximate 95% confidence intervals were calculated. In addition, to assess the measurement error and precision of the FL and CT measurements, the within-subject standard deviations were calculated using the Bland and Altman method [24]. Using this, the difference between an observed value and a participant's true value is expected to be less than two standard deviations for 95% of observations. Post hoc power calculations were carried out to evaluate whether the study was sufficiently powered, and to inform future research.
Results
The main results are shown in Table 2. No significant difference was found between the groups for FL or CT (Mann Whitney U test, FL p = 0.47, CT p = 0.91). The Hodges-Lehmann median difference between the groups was -5 ms (95% CI -32 to 8 ms) for FL, and 1 ms (95% CI -16 to 17.5 ms) for CT.
Table 2 Median (approximate 95% confidence interval) values for force generation parameters
Group FL (ms) CT (ms) Force (N) Force/Body Mass (N/kg)
Asymptomatic 62.6 (54.9–87.2) 149.8 (141.5–171.3) 5.3 (2.2–16.5) 0.08 (0.03–0.20)
OA 64.6 (54.0–109.9) 146.4 (137.1–167.3) 4.9 (1.1–18.5) 0.06 (0.01–0.25)
FL, force latency; CT, contraction time
Using the Bland and Altman method to calculate the within-subject standard deviations for FL and CT, no relationships were seen between means and standard deviations and hence this method was valid. For 95% of measurements the difference between a single observed value and a participant's true value for FL is said to be at most two standard deviations [24], i.e. 27.9 ms for asymptomatic participants and 16.4 ms for OA knee patients. For CT, these values were 29.3 ms for asymptomatic participants and 28.1 ms for OA knee patients.
To evaluate whether the study was adequately powered to detect differences between the groups in FL and CT, post hoc calculations were carried out with a significance level of 0.05. Due to the measurement error reported above, it was decided to use a value of 30 ms as the difference to be detected, as any smaller value could be less than the expected measurement error or precision of the readings. The calculations showed that 17 participants in each group were sufficient to detect a difference of 30 ms in FL at 80% power. Hence this study was sufficiently powered to detect a difference of that magnitude for this variable. However, 28 participants in each group were required to provide 80% power for CT, and this study with 17 participants had only 59% power.
The Mann-Whitney U test did not reveal any statistically significant difference between the groups in the absolute peak force of the reflex response (p = 0.72) or peak force of the reflex response when normalised to body mass (p = 0.54).
Discussion
We measured parameters of reflex force generation of the quadriceps in OA knee patients and asymptomatic participants during the patellar tendon reflex response. There were no statistically significant differences between the groups for force latency, contraction time, or force. This is an important finding as it has been reported that quadriceps muscle contraction is slower in OA knee patients than in asymptomatic individuals [9], despite the fact that this only appears to have been studied once [8], and not during reflex responses. The study by Marks et al. [8] reported that the ability to generate voluntary force quickly was impaired in the quadriceps of OA knee patients compared to young participants, but a comparison with participants of a similar age was not made. As there is evidence that a slowing of muscle contraction occurs with age [10], the difference in age between the groups in that study may have been a confounding variable. In the study here, both the OA knee and asymptomatic group comprised participants ranging from middle to old age. The fact that no significant difference was found between the groups, paired with the known significant correlation of these parameters with age [10] indicates that age-related physical changes may have more of an effect than knee joint pathology on these force generation parameters. In support of the validity of our results, our data are similar to published values, for example Burke et al. [13] reported a mean FL of 77 ms for a group of old participants, mean age 74 years.
These parameters have relevance to rehabilitation and patient care because any impairment, even only if age-related, in the time taken to initiate and then generate muscle force will increase the time taken to produce movement or stability and could impair protective reflexes [3,10]. Marks et al. [9] discuss how an improvement in the rate of force generation may reduce the rate and amount of damage to the knee joint. Strength training can increase the rate of voluntary force development [25,26], and reduce the time to peak force [27]. One of the mechanisms by which this occurs is the increase in tendon stiffness that occurs with strength training [12]. Therefore, importantly for patient care, strengthening exercises may improve contractile mechanisms that become impaired with age. These improvements may be one of the reasons why exercise is beneficial in OA knee [28] and why moderate exercise is associated with a reduced risk for OA in later life [29]. Longitudinal studies evaluating these parameters, and their associations with functional ability and pain and muscle strength would be beneficial.
However, the results of this study should be viewed with caution. The lack of a significant difference between the groups could be attributable to several explanations. These could include the small sample size, the heterogenous characteristics of the participants, other underlying conditions in the participants, and the stimulus and instrumentation used. Firstly, the results could possibly be due to a lack of study power. From our results, this study had adequate power (80%) to detect a difference of 30 ms between the groups in FL, but had only 59% power to detect this in CT. Further research with a larger sample size is therefore recommended, and this study provides data that will inform sample size calculations for future studies. To avoid making a type II error from the results observed, future studies should use a much larger total sample size, possibly necessitating a multicentre approach.
It is possible that the OA knee patients in this study could be a heterogenous group, as the general OA knee population is understood to display considerable heterogeneity [1]. No attempt was made to categorize patients in terms of aspects of OA knee classification, such as duration or level of symptoms. If this was the case in future studies, sub-group analysis could reveal whether specific sub-groups of the OA knee population do exhibit an impairment in FL and CT that could be clinically important. In addition, an assumption was made that the asymptomatic participants did not have radiographic OA. Although radiographs were not taken from the asymptomatic participants for ethical reasons, it is possible that they could have exhibited asymptomatic radiographic changes at the knee joint [30], and that this may possibly have affected the physiological responses. Future studies could also take into consideration gender and anthropometric variables, such as height and the length of the spinal reflex, that could also affect FL.
The effect of ageing on muscle function is known, but much of this effect may be indirect, i.e. it may actually be the effect of the inactivity and atrophy that may occur with ageing, and some people remain active and others do not. We did not attempt to quantify the physical activity or fitness levels of participants, or to match these between the groups. In addition, it is well known that the prevalence of other medical conditions such as neurological disease increases with age, and some participants may have had undiagnosed or sub-clinical conditions of this nature, although it is anticipated that this would have been similar in both groups.
We could possibly have obtained larger and perhaps quicker responses using a greater stimulus, as reflex response amplitude increases in a sigmoidal manner with increasing stimulus force [31,32]. The force generated in the reflex response does have an effect on FL, because as contraction force increases, electromechanical delay reduces [33,34]. In one patient, a reflex response could not be detected, and this was presumed to be due to the marked knee effusion exhibited. It is possible that a greater stimulus force could have elicited a reflex reponse in that patient. However patient comfort was an issue and reflex responses should only be compared when stimulus intensities are the same [35]. However, the repeatability of this equipment was tested and found to be constant (4.2 N) [17].
The ability to magnify or resolve traces has been shown to affect the observed electromechanical delay [36], and will probably affect values observed for all temporal parameters such as FL and CT. In evaluating the Bland and Altman precision or measurement error in FL and CT from the ten responses of each participant, we found that a single measurement could differ from a participant's true value by almost 30 ms. This may be due to the known variability in the force of the patellar tendon reflex response [23]. However there are no similar data available for FL or CT for comparison. Interestingly, this level of measurement error is far larger than the non-significant differences between the groups. Our data evaluation used modern digital methods rather than pen recordings as used in some earlier studies [13], and hence may have greater accuracy.
Finally, the open kinetic chain reflex muscle contraction tested here is a standard functional test used clinically which provides important quadriceps specific information [37]. However, this may not be representative of normal functional movement that tends to occur in the weight bearing closed kinetic chain manner [20,21]. Future studies should attempt to use both open and closed kinetic chain testing.
Much of our understanding of muscle dysfunction in OA knee remains speculative [1]. Recent studies continue to shed new light on this area, showing for example that not all OA knee patients exhibit the expected arthrogenous muscle inhibition [38], and that quadriceps strengthening may not benefit patients with malaligned knees [39]. We believe that this study adds to the evidence base and informs an area for further research that may be important and may have clinical relevance. The results of the only published paper on this topic [8] may have been affected by age differences between the two groups, but also by technological limitations. No further studies appear to have followed up this work, and this should be rectified.
Conclusion
Our experimental hypothesis in this exploratory study was that there would be a difference in FL, CT, or force between the groups during the patellar tendon reflex response. However, the results showed no statistical difference between the groups in these variables. This appears to be the first study to compare reflex FL and CT production in OA knee patients with that of asymptomatic participants. These findings provide some evidence that the time taken to generate reflex force is not impaired in OA knee patients compared to a group of similarly aged asymptomatic participants, but further research is required in this area.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Both authors participated in the design of the study, the statistical analysis, and the drafting, progress and revision of the manuscript. JD collected the data and carried out the data extraction and analysis.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors would like to thank Vicki Whittaker for statistical advice and assistance, and all the participants who took part. The study was supported by a University of Teesside School of Health and Social Care PhD Studentship.
==== Refs
Hurley MV The role of muscle weakness in the pathogenesis of osteoarthritis Rheumatic Disease Clinics of North America 1999 25 283 298 10356418 10.1016/S0889-857X(05)70068-5
Fisher NM Pendergast DR Reduced muscle function in patients with osteoarthritis Scandinavian Journal of Rehabilitation Medicine 1997 29 213 221 9428055
Hurley MV Muscle dysfunction and effective rehabilitation of knee osteoarthritis: what we know and what we need to find out Arthritis Rheum 2003 49 444 452 12794802 10.1002/art.11053
Sharma L Proprioceptive impairment in knee osteoarthritis Rheumatic Disease Clinics of North America 1999 25 299 314 10356419 10.1016/S0889-857X(05)70069-7
Radin EL Yang KH Riegger C Kish VL O'Connor JJ Relationship between lower limb dynamics and knee joint pain Journal of Orthopaedic Research 1991 9 398 405 2010844 10.1002/jor.1100090312
Jefferson RJ Collins JJ Whittle MW Radin EL O'Connor JJ The role of the quadriceps in controlling impulsive forces around heel strike Proceedings of the Institution of Mechanical Engineers 1990 204 21 28
Bland JH Cooper SM Osteoarthritis: a review of the cell biology involved and evidence for reversibility. Management rationally related to known genesis and pathophysiology Semin Arthritis Rheum 1984 14 106 133 6399624 10.1016/0049-0172(84)90002-7
Marks R Kumar S Percy J Semple J Force-time measurements of the quadriceps femoris muscles of healthy women and women with osteoarthrosis of the knee European Journal of Physical and Medical Rehabilitation 1995 5 88 92
Marks R Ungar M Ghasemmi M Electrical muscle stimulation for osteoarthritis of the knee: biological basis and systematic review New Zealand Journal of Physiotherapy 2000 28 6 20
Porter MM Vandervoort AA Lexell J Aging of human muscle: structure, function and adaptability Scand J Med Sci Sports 1995 5 129 142 7552755
Nichols TR Cope TC Abelew TA Rapid spinal mechanisms of motor coordination Exerc Sport Sci Rev 1999 27 255 284 10791019
Reeves ND Maganaris CN Narici MV Effect of strength training on human patella tendon mechanical properties of older individuals Journal of Physiology 2003 548 971 981 12626673 10.1113/jphysiol.2002.035576
Burke JR Schutten MC Koceja DM Kamen G Age-dependent effects of muscle vibration and the Jendrassik maneuver on the patellar tendon reflex response Arch Phys Med Rehabil 1996 77 600 604 8831479 10.1016/S0003-9993(96)90302-0
Koceja DM Bernacki RH Kamen G Methodology for the quantitative assessment of human crossed-spinal reflex pathways Medical and Biological Engineering and Computing 1991 29 603 606 1813757
Dixon J Howe TE Kent JR Whittaker VJ VMO-VL reflex latency difference between quadriceps components in osteoarthritic and asymptomatic knees Advances in Physiotherapy 2004 6 166 172 10.1080/14038190410020809
Altman R Asch E Bloch D Bole G Borenstein D Brandt K Christy W Cooke TD Greenwald R Hochberg M Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association Arthritis Rheum 1986 29 1039 1049 3741515
Dixon J Quadriceps activity in osteoarthritis of the knee: An electromyographic analysis 2004 , University of Teesside
Tornvall G Assessment of physical capabilities with special reference to the evaluation of maximal voluntary isometric muscle strength and maximal working capacity. Acta Physiologica Scandinavica 1963 58 Suppl 201 1 102 14013176
Edwards RH Young A Hosking GP Jones DA Human skeletal muscle function: description of tests and normal values Clin Sci Mol Med 1977 52 283 290 844260
Callaghan MJ Oldham JA The role of quadriceps exercise in the treatment of patellofemoral pain syndrome Sports Medicine 1996 21 384 391 8724204
Stensdotter AK Hodges PW Mellor R Sundelin G Hager-Ross C Quadriceps activation in closed and in open kinetic chain exercise Med Sci Sports Exerc 2003 35 2043 2047 14652500
Rothwell JC Day BL Berardelli A Marsden CD Habituation and conditioning of the human long latency stretch reflex Exp Brain Res 1986 63 197 204 3732444 10.1007/BF00235664
Toft E Sinkjaer T Rasmussen A Stretch reflex variation in the relaxed and the pre-activated quadriceps muscle of normal humans Acta Neurologica Scandinavica 1991 84 311 315 1772004
Bland JM Altman DG Statistical methods for assessing agreement between two methods of clinical measurement Lancet 1986 1 307 310 2868172
Kubo K Kanehisa H Ito M Fukunaga T Effects of isometric training on the elasticity of human tendon structures in vivo Journal of Applied Physiology 2001 91 26 32 11408409
Häkkinen K Alen M Komi PV Changes in isometric force- and relaxation-time, electromyographic and muscle fibre characteristics of human skeletal muscle during strength training and detraining Acta Physiologica Scandinavica 1985 125 573 585 4091001
Narici MV Hoppeler H Kayser B Landoni L Claassen H Gavardi C Conti M Cerretelli P Human quadriceps cross-sectional area, torque and neural activation during 6 months strength training Acta Physiologica Scandinavica 1996 157 175 186 8800357 10.1046/j.1365-201X.1996.483230000.x
McCarthy CJ Oldham JA The effectiveness of exercise in the treatment of osteoarthritic knees: a critical review Physical Therapy Reviews 1999 4 241 250
Manninen P Riihimaki H Heliovaara M Suomalainen O Physical exercise and risk of severe knee osteoarthritis requiring arthroplasty Rheumatology 2001 40 432 437 11312383 10.1093/rheumatology/40.4.432
Felson DT Naimark A Anderson J Kazis L Castelli W Meenan RF The prevalence of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study Arthritis Rheum 1987 30 914 918 3632732
Zhang LQ Huang H Sliwa JA Rymer WZ System identification of tendon reflex dynamics IEEE Transactions on Rehabilitation Engineering 1999 7 193 203 10391590 10.1109/86.769410
Stam J Tan KM Tendon reflex variability and method of stimulation Electroencephalogr Clin Neurophysiol 1987 67 463 467 2444415 10.1016/0013-4694(87)90010-1
Viitasalo JT Komi PV Interrelationships between electromyographic, mechanical, muscle structure and reflex time measurements in man Acta Physiologica Scandinavica 1981 111 97 103 7223456
Vint PF McLean SP Harron GM Electromechanical delay in isometric actions initiated from nonresting levels Med Sci Sports Exer 2001 33 978 983 10.1097/00005768-200106000-00018
Pagliaro P Zamparo P Quantitative evaluation of the stretch reflex before and after hydro kinesy therapy in patients affected by spastic paresis Journal of Electromyography and Kinesiology 1999 9 141 148 10098714 10.1016/S1050-6411(98)00045-5
Corcos DM Gottlieb GL Latash ML Almeida GL Agarwal GC Electromechanical delay: An experimental artifact Journal of Electromyography and Kinesiology 1992 2 59 68 10.1016/1050-6411(92)90017-D
Frijns CJ Laman DM van Duijn MA van Duijn H Normal values of patellar and ankle tendon reflex latencies Clin Neurol Neurosurg 1997 99 31 36 9107465 10.1016/S0303-8467(96)00593-8
Lewek MD Rudolph KS Snyder-Mackler L Quadriceps femoris muscle weakness and activation failure in patients with symptomatic knee osteoarthritis Journal of Orthopaedic Research 2004 22 110 115 14656668 10.1016/S0736-0266(03)00154-2
Sharma L Dunlop D Cahue S Song J Hayes KW Quadriceps strength and osteoarthritis progression in malaligned and lax knees Annals of Internal Medicine 2003 138 613 619 12693882
|
16137326
|
PMC1232853
|
CC BY
|
2021-01-04 16:32:04
|
no
|
BMC Musculoskelet Disord. 2005 Sep 1; 6:46
|
utf-8
|
BMC Musculoskelet Disord
| 2,005 |
10.1186/1471-2474-6-46
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-821609295810.1186/1471-2458-5-82Research ArticleAn evaluation of a morphine public health programme for cancer and AIDS pain relief in Sub-Saharan Africa Logie Dorothy E [email protected] Richard [email protected] Cheviot View, Bowden, Melrose, UK, TD6 OST2 Department of Palliative Care & Policy, Guy's King's & St Thomas' School of Medicine King's College London UK2005 10 8 2005 5 82 82 20 4 2005 10 8 2005 Copyright © 2005 Logie and Harding; licensee BioMed Central Ltd.2005Logie and Harding; 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
Despite growing HIV and cancer prevalence in Sub-Saharan Africa, and WHO advocacy for a public health approach to palliative care provision, opioid availability is severely limited. Uganda has achieved a morphine roll-out programme in partnership with the Ministry of Health. This study aimed to evaluate that programme by identifying challenges to implementation that may inform replication.
Methods
A multi-methods protocol appraised morphine regulation, storage, prescribing, and consumption in three phases: key informant interviews throughout the opioid supply chain, and direct observation and audit of clinical practice.
Results
Regulation had achieved its goal of preventing misuse and leakage from the supply chain. However, the Government felt that relaxation of regulation was now appropriate. Confusion and complexity in storage and authorisation rules led to discontinuation of opioid pain management at the patient level and also wasted service time in trying to obtain supplies to which they were entitled. Continued neglect to prescribe among clinicians and public fear of opioids led to under prescribing, and clinical skills showed some evidence of need for improvement with respect to physical assessment and follow-up.
Conclusion
The Ugandan programme offers a successful model for both advocacy and Governmental support in achieving opioid roll-out across health districts. Despite initial concerns, abuse of opioids has not been evident. Further work is required to ensure that available supplies of opioids are prescribed to those in need, and that clinical standards are met. However, the programme for roll-out has proved a useful model to expand opioid availability as the first step in improving patient care, and may prove a useful template for other Sub-Saharan African countries.
==== Body
Background
Most, if not all, pain due to cancer and AIDS could be relieved if we implemented existing medical knowledge [1]. This does not happen for two thirds of the world's population for several reasons. Most live in countries with weak public health infrastructure and, despite WHO encouragement for a public health approach to pain management and palliative care [2], a lack of legislative and policy support exists. Despite an HIV/AIDS epidemic which kills 3 million each year, inadequate availability of pain medication, especially of opioids, is evident in that only eleven out of forty seven African countries use morphine for chronic pain and, of these eleven, the amount consumed is tiny [3]. South Africa is highest user on the continent consuming 265 daily defined doses (DDD) while Namibia uses 97 and Uganda 9 DDDs, and eighty-six percent of the world's morphine is still used by the 20 richest countries [4]. Sadly, although morphine is a safe drug which can be used with anti-retroviral therapy, chemotherapy and traditional medicines, it is under-used largely due to professional fears.
Uganda is the first African country to follow the WHO guidelines, and, despite limited resources and a population which is 90% rural, has prioritised palliative care under Essential Clinical Services in the National Health Plan (2001–2005). It has made oral morphine freely available to those districts that have specialist palliative care nurses or clinical officers and has promoted morphine use down to village level. It has passed laws to allow nurse prescribing of morphine [5], an essential step as doctors are scarce in rural areas. There are no limits to the number of days, nor dose, which doctors or trained nurses and clinical officers can prescribe within the hospice setting, though only the weaker strength (5 mgs per 5 mls) was available to community nurses outside the Hospice setting.
Since 1993, Hospice Africa Uganda (HAU), a non-governmental organisation, has pioneered three community- based palliative care programmes in rural and urban localities. However, coverage has been comparatively low. In 1998, following several years of lobbying, the Ministry of Health invited hospice staff to be technical experts in a pilot study in 15 (out of 56) rural and urban districts to assess the feasibility and safety of using morphine for chronic pain in the community. This included cancer pain, pain from HIV/AIDS disease, and pain from sickle cell crisis. In 2002–03, the 15 districts, including mission hospitals, underwent extensive initial training involving local dignitaries, police, and senior health officials. The programme is still in the early stages and has faced challenges in funding, staffing and supervision. This paper reports an evaluation of the Ugandan morphine access programme, which aimed to appraise the processes of morphine supply, to assess clinical practice on prescribing and pain control, and investigate patient satisfaction with symptom control and costs. This initial outline of a working model of morphine access, achieved through policy, legislative and clinical efforts, was developed as an example for the continent. The strengths and blockages in the system of morphine delivery are described.
Methods
Setting
The study was undertaken in 2003 across rural and urban Ugandan hospice sites, two of the pilot districts Government District Hospitals, the Ministry of Health, and 2 home-based care NGO programmes. Ethics permission was obtained from the Ethics Committee of Liverpool School of Tropical Medicine and from Hospice Africa Uganda. Permission from participants was obtained by reading a standard permission sheet which explained about the survey, that it was confidential, voluntary, and would report anonymously. This was translated into local languages
Procedure
The evaluation protocol employed mixed data collection methods from stakeholders throughout the supply chain.
Phase 1: semi-structured interviews among three populations. Firstly, clinicians prescribing morphine (n = 16); secondly, patients accessing morphine (n = 10); thirdly, key informants (n = 16) including senior clinical and Governmental staff. Interviews were voluntary and confidentiality was assured. Translation for patients was provided. Interviews were tape recorded, transcribed verbatim and thematically coded, and themes developed adopting framework analysis [6] using Winmax software.
Phase 2: direct observation of morphine from entry to the country to a pharmaceutical NGO and on to distribution through hospital wards and patients' homes. Pharmacy, ward and nursing records were checked against legal requirements [7].
Phase 3: two quality of clinical care audits across two hospice sites, urban and rural. The first audit observed domiciliary nurse visits prescribing oral morphine (n = 21), the second retrospectively examined prescribing standards in randomly-sampled clinical notes (n = 50). Criteria were established using local prescribing guidelines and clinical best-practice literature [8-10]. The random files sample was generated by selecting every 5th deceased patient file for deaths recorded during a 12-month period 2002–2003.
Results
The data on barriers and blockages in morphine supply are represented in the model in Figure 1.
Figure 1 Morphine supply and blockages within the Ugandan regulatory framework. Blockages in morphine flow: key. District Hospital has to obtain permission from MOH for every order and take this to JMS. National Medical Stores currently re-structuring and lacks capacity and drug inspectors. Mission hospitals and NGOs have to obtain morphine via District Hospital pharmacy and depend on the co-operation of the pharmacist/dispenser. Pharmacy may be slow in re-ordering. Local health centre/hospice level, drug cupboard may not be adequate, record books may be missing, there may be staff turn over resulting in loss of know-how.
Morphine regulations
Opioids are monitored by the International Narcotics Control Board which aims to promote Government compliance with international drug treaties and assist them in this effort. The National Drug Regulatory Authority operates through the Ugandan Ministry of Health, which estimates Uganda's needs each year and writes laws which cover the production of Class A drugs (WHO schedule 2) manufacture, distribution, use, and registration of all handlers. Uganda has produced model regulation guidelines [11] which are available as a model for other countries. Uganda uses morphine sulphate made up generically in-country from powder to liquid and costing between 1.8–5.6 US$ per patient per month [12]. It is stable in liquid form without refrigeration for up to three months making it especially suitable for community use.
When the scaling up of morphine use to the community was first discussed, there was suspicion from politicians, police and senior doctors who feared Uganda would become "a smugglers paradise". Therefore, the balance between encouraging use and preventing misuse was weighted towards safety. As there has not been a single case of narcotic abuse over the twelve years of hospice and three years of government use, there was a reported consensus view among Governmental and regulatory bodies that initial caution could be relaxed. This may have a significant patient impact.
"Nurses fear blame if they cannot account for all their supplies of morphine. A 24-year old man with Kaposi's sarcoma was found at home in a remote area, in great pain, with gross liver distension and dehydration. Worried about the inadequacy of pain relief it was suggested that an extra bottle (250 mls) might be left in case of breakthrough pain. The nurse prescriber was reluctant. She agreed he might soon die but argued that she might not then be able to retrieve the unused supply. The man was not seen again and was reported dead a few days later he died with prolonged convulsions. The nurse said: "Some of our clients have passed away and we can't find out where the morphine is. The previous doctor was very strict and he would check our books".
Regulation involves checking authorisation, complex book-keeping, drug storage, and occasionally caused blockages in the morphine flow (see Figure 1). Confusion arose over accreditation of NGO health professionals who had to obtain supplies through the district hospitals, as opposed to through national medical stores. One palliative care trained doctor from an NGO was refused morphine as the district hospital wanted written evidence of authority, without providing the necessary clarity about how this could be obtained. As a result a major NGO treating HIV/AIDS was without morphine. Double authorisation meant that district hospital teams had to visit two locations (MOH and medical stores) in Kampala for renewed supplies. Staff experienced difficulty in finding the appropriate person to authorise the prescription and one nurse waited 6 hours for a signature. Another had to make two journeys from up-country occupying a scarce hospital vehicle.
The regulations dictate that morphine must be double locked in a secure cupboard attached to the wall. Often pharmacy rooms were inadequate or cupboards too small to hold 500 mls bottles. Obtaining bigger cupboards was a major challenge because of different funding pathways. Drug record books could be difficult to obtain and some clinical staff found them complex to complete.
"We started to introduce morphine. They (our patients) need it. The problem is we don't have a locked cupboard......and the pharmacy is so small. It sounds easy (to put right) but they have a different administration (from the health authorities)". NGO Prescriber
"Often the pharmacist/dispenser in the hospital knew how to fill out the books however there was lack of consistency in how it was undertaken between different hospitals". Hospice Staff
New laws have been introduced allowing nurse prescribing but, until recently, doctors had to counter-sign prescriptions. The ease of achieving this depended on the doctors' attitude and much nursing time was spent chasing signatures. One nurse said that some health professionals do not see the terminally ill as a priority.
Under-use of morphine
The patients were grateful to have been prescribed morphine which reportedly improved their quality of life, felt that it prolonged life, and had few side effects (at 5 mgs per 5 mls, the only strength available at District level). One patient said "I cannot imagine being in that state (of pain) again". However professional fears about morphine proved difficult to shift and, despite undergoing a 5-day training programme, one district hospital returned the first supplies unused.
"There are 10 doctors in the hospital and only 3 are willing (to sign prescriptions). The others you have to keep begging. I visit one ward with cancer and HIV patients and there I have to keep asking him to write but he doesn't write. He doesn't take it seriously or he doesn't think he should take it from a nurse who is lower (than he is)". Nurse Prescriber
Some nurses blamed doctors for perpetuating fears that morphine hastened death while others thought public opinion responsible. In one rural area local radio was used to re-assure people that the palliative care team "do not kill".
"In Hoima they say "when you go to that doctor you are going to die(because they give strong medicine) so we are starting to plan a radio programme maybe quarterly or monthly to talk about palliative care and try to wipe away that image that we kill". Nurse Prescriber
Lack of essential drug and human resource shortages
Morphine should be used with other analgesics according to the WHO analgesic ladder. Use was curtailed if the district hospital ran out of simple analgesics or laxatives. Patients were asked to buy such supplies and many could not afford to do so.
Staff shortages also hindered availability. Although the HAU is well-staffed, government hospitals had severe nurse shortages. The tertiary referral hospital in Kampala with 33 wards had only one palliative care nurse to prescribe and monitor morphine (a second was in training). She covered only two wards, cancer and gynaecology, and palliative care was unavailable to many HIV/AIDS patients.
With only 200 pharmacists in the country, and 300 pharmaceutical assistants, fewer than 10 Districts in Uganda have a trained pharmacist. Nurses felt that having a pharmacist with a positive attitude to morphine use was key to obtaining uninterrupted supplies. The serious shortage of drug inspectors also weakened the functioning of the system.
Quality of care
Quality was measured in two audits (phase 2) from rural and urban hospice sites (see Table 1). Clinical care standards in trained nurses were high although, in 7 cases, the clinical examination was considered by the observer to be incomplete (no torch examination of the mouth, no abdominal or rectal examination in the presence of constipation, no examination of the spine in suspected vertebral collapse). Validated pain measurement to titrate morphine was under-used in both audits with a reduction in completion of pain scores across successive visits. Follow up after starting morphine is important to assess adequacy of pain relief and potential side effects. Thirty-eight patients (76%) were not re-visited within the recommended 3 days, and in 25 cases (50%) the follow-up was 5 days or longer. The method of procuring further morphine was not clear in 74% of notes (n = 37) although the information may have been verbally delivered. However, prescription details were clearly written in 84% files (n = 42).
Table 1 Audit data: observation of nurse consultations in patients' homes (n = 21) and file reviews
Yes Frequency (Percent) No Frequency (Percent) Missing
Pain measurement questions
Was there a measure of quality of life (sleep and mobility)? 17 (81%) 4 (19%) 0
Was the pain scale (VAS) used to monitor pain? 5 (23.8%) 16 (76.2%) 0
Was duration of the pain noted? 18 (85.7%) 3 (14%) 0
Consultation quality indicators
Were patient's views on morphine sought? 6 (28.6%) 15 (71.4%) 1
Were instructions clearly given? 19 (90.5%) 2 (9.5%) 1
Were side effects discussed? 14 (66.7%) 7 (33.3%)
Was a laxative prescribed? 14 (66.7%) 5 (23.8%) 2 (9.5%)
Was the patient fully examined? 13 (61.9%) 7 (33.3%) 1
Was a follow up date arranged? 21 (100%) 0 0
Were details of the prescription written in notes? 20 (95.2%) 0 1 (4.8%)
Financial costs of illness are only one aspect. At home or in hospital, relatives have to care for the sick and this may result in the principal bread winner giving up work.
Identifying patients and following them up
The logistics of identifying and following-up patients in rural areas are complex. With only one dedicated vehicle or a communal hospital vehicle, co-operation with community volunteers was invaluable. One rural hospice team said they could not function without them. Such liaisons ran into trouble for lack of funds to pay even expenses, let alone "incentives". If a patient lived far away, the volunteer or relative were asked to return for repeat drugs and report progress, but this incurred expense and time. Frequently, patients presented late and were not reviewed after the first morphine prescription.
Costs to patient
Although user fees were officially banned in Uganda in 2002, charges for treatment are still in place as financial pressures are passed on. The Hospice charged patients 5,000 shillings (£2.70) per week for all care including drugs, which only one third could afford. Even when morphine is free, patients expressed distress at the costs of transport to collect it. Some patients move nearer to clinics in order to access a continuous supply of pain relieving drugs, subsequently ending their lives far from their children and families.
"If I am in Congo there is nothing there (no drugs). For 2 months in Congo I was without drugs. I spread the dose (2 week's worth) out for 2 months. I'd like to go home but can't because of the tablets. That is why I am here". Patient
"I stay about 30 miles away but I come to stay in my son's place in Kampala. The Hospice picks me up with their transport (weekly). It costs 5,000 shillings (£2.75) every week for the hospice. But the journey home costs 6,000 shillings (£3) to my village. If I can find transport money I go home. Usually once a month to look after my garden". Patient
Discussion
From this initial evaluation of this innovative multi-agency morphine roll-out programme, we wish to disseminate 5 key lessons learned from the Ugandan experience.
First, regulation. As demand for morphine expands, adequate supply at national level must be assured by forward planning. The two functions, opioid availability and prevention of narcotic abuse, should be separate as too many legal obstacles for busy health professionals are a disincentive to use. No country has recorded diversion of therapeutic oral morphine to illegal use and many experts believe that stiff controls could be relaxed, at least at the prescriber-patient level [13]. There is no reason, if district orders match district estimates, that morphine cannot be obtained in the same way as other drugs.
Second, on-going support. It is important that the palliative care programme is owned by district health authorities. Although initial district training is costly in terms of nurse time, transportation and per diems, new money is now becoming available through the Global Fund and the President's Emergency Plan For AIDS Relief which includes funding for palliative care. Monitoring the quality of the service is important as palliative care nurses work in isolated environments. The Palliative Care Association of Uganda, under the umbrella of the new pan-African Palliative Care Association (APCA), aims to help develop and promote standards.
Third, nurse training. In Uganda it takes 9 months to become a specialist palliative care nurse (and be allowed to prescribe). Since 1993, 600 health professionals have been trained. Some question the length of training in view of the huge unmet need. But, if shortened, nurses' ability to diagnose pain accurately and prescribe effectively might be jeopardised. The Director of HAU suggests that the Ugandan training programme could be adapted to suit different health infrastructures and different types of nurses and proposes that a group of nurses and one doctor could be trained per region, who could then train others.
Fourth, the use of community volunteers. Community based volunteers are essential but lack of funding, regulation, training and support are concerns [14,15]. In Zambia, it was shown that transport costs alone accounted for 33% of home-carer costs [16]. HAU believes that without a proper palliative care component home care can be equated to home neglect [17]. The answer might be for clinical staff to lead and supervise a team of carers including traditional healers.
Fifth, the danger of health care costs displacing other essentials such as food. Forty-two percent of monthly Ugandan household expenditure is spent on food [18], and illness diverts money from this. A WHO sponsored survey in Botswana, Ethiopia, Tanzania, Zimbabwe and Uganda [19] identified practical needs of the dying and found loss of income was a huge problem for two thirds of the patients. Multi-sectoral links between a palliative care service and organisations supplying food and poverty reduction are essential to alleviate health-related family poverty.
Conclusion
The magnitude of unnecessary suffering in countries like Uganda is so huge that pain relief should be as important a public health priority as anti-retroviral therapy, and the two closely linked. As money becomes available from the Global Fund, and as 15% of PEPFAR funding has been targeted for palliative care, this will help support other national palliative care programmes. But careful planning, education and monitoring are necessary. The Ugandan experience may be seen as a result of concerted lobbying and policy wok through HAU, which is likely to be necessary in other countries that wish to replicate their success. A recent review of palliative care in Sub-Saharan Africa found a chronic lack of African-relevant evidence on outcomes in palliative care [20,21] despite a wealth of practitioner experience. It is hoped that lessons learnt from Uganda will add to the knowledge and encourage other countries to adopt similar programmes to relieve the epidemic of silent suffering.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
DL designed the study and undertook data collection. DL and RH analysed the data and drafted the manuscript. Both authors read and approved the final manuscript.
Table 2 Myths and barriers to prescribing morphine
MYTHS BARRIERS
Professional fears about safety of morphine and addiction Logistic supply chain and transport inadequacies
Public fears that morphine expedites death Lack of pharmacists and pharmacy support
Perceived difficulties in predicting national requirements Lack of trained palliative care staff
Fears about illegal diversion Over regulation, legal barriers and complex regulations
Fear of inaccurate diagnosis Difficulties finding patients in need of care and following them up
Perception that all therapeutic morphine prescribed needs to be accounted for Lack of hospital/palliative care vehicles
Low priority given by medical staff to the dying Palliative care nurse training lengthy
Cost of complimentary drugs e.g. laxatives or Step 1 and 2 analgesics
Perceptions that laws governing therapeutic morphine are difficult to change Morphine storage difficulties
Health staff overwhelmed
Low priority by hospitals to palliative care
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Thanks to the staff in the International Health Research Group at Liverpool School of Tropical Medicine and to Dr Jack Jagwe, Dr Anne Merriman, Dr Ekira Kikule, and Diana Opio, Education Department Hospice Africa Uganda for their help.
DL is grateful to The Kenneth Newell Bursary for providing financial assistance.
==== Refs
World Health Organisation Cancer pain relief and palliative care 1990 Geneva: WHO
World Health Organisation Cancer pain relief; with a guide to opiate availability 1996 Geneva: WHO
International Narcotics Control Board Use of essential narcotic drugs to treat pain is inadequate, especially in developing countries 2004
Spencer M Rajagopal M, Mazza D, Lipman A Pain relief in Thailand Pain and palliative care in the developing world 2003 New York: The Haworth Medical Press
Ministry of Health (Uganda) Health Sector Plan 2003 2003 Kampala
Bryman A Burgess RG Analyzing qualitative data 1994 London: Routledge
Ministry of Health (Uganda) Handling Class A Drugs 2001 Kampala
Doyle D Hanks GW Oxford Textbook of Palliative care medicine 1998 2 Oxford: Oxford University Press
Merriman A Benton TF Pain and symptom control in the cancer and/or AIDS patient in Uganda and other countries 2002 3 Kampala: Hospice Africa Uganda
Twycross RG Introducing palliative care 1995 Oxford: Radcliffe Press
Ministry of Health (Uganda) Handling Class A Drugs 2001 Kampala: Ministry of Health
Stjernsward J Rajagopal M, Mazza D, Lipman A Instituting palliative care in developing countries – an urgently needed and achievable goal Pain and palliative care in the developing world 2003 New York: The Haworth Medical Press
Rajagopal M Joranson D Gilson A Medical use, misuse, and diversion of opioids in India Lancet 2001 358 139 143 11463435 10.1016/S0140-6736(01)05322-3
Nsutebu EF Walley JD Mataka E Simon CF Scaling-up HIV/AIDS and TB home based care: lessons from Zambia Health Policy Plan 2001 16 240 7 11527864 10.1093/heapol/16.3.240
Murray SM Grant E Grant A Kendall M Dying from cancer in developed and developing countries: lessons from two qualitative interview studies of patients and their carers BMJ 2003 326 368 371 12586671 10.1136/bmj.326.7385.368
Essex M Mboup S AIDS in Africa 2002 2 New York: Kluwer Academic
Kikule E A good death in Uganda: survey of needs for palliative care for terminally ill people in urban areas BMJ 2003 327 192 4 12881259 10.1136/bmj.327.7408.192
Ministry of Finance (Uganda) Planning and Economic Development (MoFPED) Poverty Status Report 2002 Kampala: Ministry of Finance
Sepulveda C Habiyambere V Amandua J Borok M Kikule E Mudanga B Quality care at the end of life in Africa BMJ 2003 327 209 213 12881267 10.1136/bmj.327.7408.209
Harding R Higginson IJ Palliative care in Sub-Saharan Africa: an appraisal 2004 London, UK, Diana, Princess of Wales Memorial Fund
Harding R Higginson IJ Palliative Care in Sub-Saharan Africa: an appraisal of reported activities, evidence and opportunities Lancet 2005 365 1971 1977 15936427 10.1016/S0140-6736(05)66666-4
|
16092958
|
PMC1232854
|
CC BY
|
2021-01-04 16:28:57
|
no
|
BMC Public Health. 2005 Aug 10; 5:82
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-82
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-891612022510.1186/1471-2458-5-89Research ArticlePatterns of HIV prevalence among injecting drug users in the cross-border area of Lang Son Province, Vietnam, and Ning Ming County, Guangxi Province, China Des Jarlais Don C [email protected] Patrick [email protected] Patricia [email protected] Ryan [email protected] Wei [email protected] Doan [email protected] Yi [email protected] Tran V [email protected] Meng [email protected] Ly K [email protected] Nguyen D [email protected] Kieu T [email protected] Theodore M [email protected] Baron Edmond de Rothschild Chemical Dependency Institute, Beth Israel Medical Center, New York City, USA2 Abt Associates, Inc., Cambridge, USA3 Guangxi Ctr. for HIV/AIDS and Prevention and Control, Nanning, China4 Hanoi, Vietnam5 Family Health International, Hanoi, Vietnam6 Ning Ming County Health Department, Ning Ming City, Guangxi, China7 Lang Son Provincial Health Service, Lang Son, Vietnam8 General Department of Preventive Medicine and HIV/AIDS Control, Ministry of Health, Hanoi, Vietnam9 Hanoi, Vietnam2005 24 8 2005 5 89 89 29 10 2004 24 8 2005 Copyright © 2005 Des Jarlais et al; licensee BioMed Central Ltd.2005Des Jarlais 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 assess patterns of injecting drug use and HIV prevalence among injecting drug users (IDUs) in an international border area along a major heroin trans-shipment route.
Methods
Cross-sectional surveys of IDUs in 5 sites in Lang Son Province, Vietnam (n = 348) and 3 sites in Ning Ming County, Guangxi Province, China (n = 308). Respondents were recruited through peer referral ("snowball") methods in both countries, and also from officially recorded lists of IDUs in Vietnam. A risk behavior questionnaire was administered and HIV counseling and testing conducted.
Results
Participants in both countries were largely male, in their 20s, and unmarried. A majority of subjects in both countries were members of ethnic minority groups. There were strong geographic gradients for length of drug injecting and for HIV seroprevalence. Both mean years injecting and HIV seroprevalence declined from the Vietnamese site farthest from the border to the Chinese site farthest from the border. 10.6% of participants in China and 24.5% of participants in Vietnam reported crossing the international border in the 6 months prior to interview. Crossing the border by IDUs was associated with (1) distance from the border, (2) being a member of an ethnic minority group, and (3) being HIV seropositive among Chinese participants.
Conclusion
Reducing the international spread of HIV among IDUs will require programs at the global, regional, national, and "local cross border" levels. At the local cross border level, the programs should be coordinated on both sides of the border and on a sufficient scale that IDUs will be able to readily obtain clean injection equipment on the other side of the border as well as in their country of residence.
==== Body
Background
Both injecting drug use and HIV among injecting drug users (IDUs) have become major international public health problems. HIV infection has been reported among IDUs in over 100 countries [1]. Travel by IDUs, particularly along drug distribution routes, appears to be a major mechanism for the spread of both injecting drug use and HIV among IDUs. HIV spread north [2] and south [3] from New York City along the East Coast in the U.S. Stimson [4], and Beyrer and colleagues [5] have reconstructed the spread of HIV among IDUs in South East Asia. Beyrer et al. used molecular epidemiology (mapping the different subtypes of HIV) in their reconstruction. A recent study by Kato and colleagues (2001) found patterns of HIV genetic subtyping consistent with cross-border transmission either from Vietnam to China or from China to Vietnam.
While these regional and country level analyses have great value in understanding the worldwide spread of HIV among IDUs, they have important limitations with respect to HIV prevention efforts. Reducing HIV spread by attempting to disrupt regional and country drug distribution routes may have the unintended consequence of displacing distribution to new routes, leading to additional spread of injecting drug use and HIV among IDUs.
Successful prevention efforts will be greatly facilitated by more detailed understanding of the spread of injecting drug use and transmission of HIV among IDUs within smaller geographic areas. Understanding of HIV transmission across international borders is particularly important, as few HIV prevention programs are coordinated across such borders. We present here data on injecting drug use and HIV among IDUs in the adjacent border provinces of Lang Son, Vietnam and Guangxi, China. HIV among IDUs was noted in this area in 1996 [6] and since then there has been substantial transmission among IDUs in both provinces. The present situation shows a clear geographic pattern, with the potential for additional spread across the border between the provinces and within each of the provinces.
The data reported here were collected as part of baseline surveys of IDUs conducted before implementation of a cross-border HIV prevention intervention in Lang Son Province, Vietnam and-Ning Ming County, Guangxi Province, Vietnam [7]. Figure 1 shows a map of the area, with the project sites – in Lang Son Province and in Ning Ming County.
Figure 1 Geographic Setting and map of project sites.
There is considerable official and entirely legal movement of commercial goods across the border in both directions. There are also semi-official and informal crossing points and pathways through the hills that permit local residents to cross the border with little or no regulation. Crossing the border is a regular aspect of life for many people who live near the border, for example, to attend market days in the larger villages. Drug dealers cross the border to sell drugs and drug users also cross to obtain drugs of higher purity and at better prices. The frequency of border crossing can vary, influenced by current price and purity of heroin, the ebb and flow of law enforcement activity and, factors such as the outbreak of severe acute respiratory syndrome (SARS) in China.
There is also substantial trade and migratory employment in the region. Lang Son City, Aidian, and Puzhai (near Pingxiang) are bustling centers of legitimate cross-border trade, as well as drug trafficking and sex work. Many people cross the border daily and seasonally to find work and many are employed as porters in the cross-border trade. The region is home to many ethnic minority groups (e.g., Zhuang, Tay, Nung), some of whom live on both sides of the border (for example, the Zhuang in China and the Tay in Vietnam are the same ethnic group but are known by different names in the two countries). There is frequent intermarriage across the border, although this might be illegal and might lead to loss of nationality. Kinship ties, like migratory employment and trade, result in additional cross-border movement.
Methods
Data collection methods for the IDU surveys reported on here were essentially parallel in Ning Ming County and Lang Son Province, with some variation in the community-based subject recruitment strategies used. The availability of large known drug use gathering places and of officially registered IDUs in Lang Son permitted greater use of probability-based methods in Vietnam, while there was more reliance on peer recruitment in China.
Subject recruitment – China
In Ning Ming County, a modified snowball/peer recruitment technique was used. The project peer educators sent recruiting letters to IDUs they knew personally, inviting them to come to a project center and participate in the survey. The IDUs who came to project centers for interviews were encouraged to recruit 2–3 additional participants. The research participants received 20 Chinese yuan (approximately $2.50) for the interview, 5 yuan for each additional male respondent recruited, and 10 yuan for each additional woman respondent recruited. The eligibility criteria were a minimum of 18 years of age and recent (in the past 6 months) drug injection.
Subject recruitment – Vietnam
Approximately one-half of the sample was based on individuals initially selected from the lists of known IDUs in the project sites. The other half was based on participants initially selected from IDUs present at gathering or shooting places mapped by project staff as part of the initial project implementation. For the half of the sample based initially on registered lists, 10 clusters of 25 individuals each were selected by probability proportional to size (PPS) from the lists of IDUs in each commune. Then four IDUs were picked at random from each selected cluster and these referred others until the quota for the commune was reached.
For the portion of the sample selected initially at IDU gathering or shooting places, sample quotas for these places were determined by PPS based on the numbers of individuals observed at these places during the mapping phase. The interview team then revisited the selected places and chose four individuals at random from among those present at each place at that time (who were not necessarily those present during the mapping phase).
The Vietnamese participants were paid 30,000 Vietnamese dong (approximately US$2) for participating in the interview and HIV test.
Informed consent
In Vietnam, an oral informed consent was obtained for participation in the study, with the interviewer certifying that oral consent had been obtained. This procedure was requested by the Institutional Review Board of the National AIDS Standing Bureau in order to provide more assurance of confidentiality to prospective participants. In China, standard signed informed consents were obtained from all participants. Unique codes were constructed for each participant based on numeric date of birth and several letters representing, for example, the first letter of the mother's family name. (Construction of the record number was slightly different in the two countries.) The objective was to have a unique identifier that the participant could readily reconstruct if he or she lost the project participation card.
Questionnaire
A structured instrument was used for the interviews, based on version 2b of the questionnaire being used in the World Health Organization's Drug Injection Study, Phase II [8]. Trained interviewers, primarily staff of the local health departments, conducted the interviews. The questionnaire covered demographics, drug use, injection and sexual risk behavior, HIV testing history, HIV and hepatitis knowledge, and cross-border travel patterns. A question on the number of times the subject had crossed the border in the 6 months prior to the interview was included.
There are many factors which could influence the "ease/difficulty" in crossing an international border, including distance to the border, cost of transportation, time needed to reach the border, and the need to have official papers for crossing. It was not practical to measure all such factors. Instead, used simple physical distance (in kilometers) to the nearest border point. This gave five distance categories in Vietnam and three distance categories in China.
The baseline survey was conducted in July 2002 in Vietnam and between July and September 2002 in China.
HIV testing
The survey included HIV antibody testing. Participants were given pre-test counseling and post-test counseling at local health centers. Blood was drawn at the time of the interviews by trained phlebotomists from local health departments. Participants were given a card with their unique identifier and returned to the local health center to receive their test results using this identification number. Indeed, they could only retrieve their results by using this number since blood samples were not otherwise labeled. In China, testing was by double ELISA (Vironostika HIV- Uni-form, Organon (Holland)) with confirmation of initial HIV-positive results by Western Blot (Genelabs Diagnostics). All testing was conducted at the laboratory of the Guangxi Center for HIV/AIDS Prevention and Control in Nanning. In Vietnam, testing was performed at the laboratory of the Lang Son Provincial Health Services using the Serodia SFD screening test (Biorad {France}) and double ELISA (Genescreen, Biorad (France); Vironostika, Organon [9]). This is the official protocol of the Ministry of Health in Vietnam and the Lang Son laboratory is authorized to provide HIV testing according to this protocol by the Ministry of Health.
Data analysis
Data were entered and data sets were prepared in EpiInfo, version 6.04 by staff of the Guangxi Center for HIV/AIDS Prevention and Control and the National AIDS Standing Bureau of Vietnam (which has since been merged into the General Department of Preventive Medicine and HIV/AIDS Control of the Ministry of Health). The data were analyzed at Beth Israel Medical Center and Abt Associates Inc. using the Statistical Analysis System, Version 8.2 (SAS, Inc., Cary, NC). Chi square tests were used to assess bivariate relationships among categorical variables. Multi-collinearity problems precluded multivariate analyses.
Ethical review
The study was reviewed and approved by the institutional review boards (IRBs) of the following institutions: Guangxi Center for HIV/AIDS Prevention and Control, the National AIDS Standing Bureau of Vietnam, Abt Associates Inc., and Beth Israel Medical Center.
Results
Table 1 shows selected demographic characteristics of the IDU subjects recruited in China and Vietnam. In both provinces, the subjects were primarily young males who had never been married. Although there are known to be female IDUs on both sides of the border, the project has had difficulty inducing women to participate in the interventions and in recruiting them for the surveys. Over two-thirds of the subjects in China and one-half in Vietnam belonged to ethnic minority groups (primarily Zhuang in China and Tay and Nung in Vietnam). The ethnic minority subjects tended to live closer to the border. Table 2 and 3 shows the percentages of ethnic minority subjects in the different sites by distance to the border (in kilometers).
Table 1 Characteristics of the Baseline IDU Samples
Characteristic China (n = 294) Vietnam (n = 348)
Male 90% 99%
Age 21–30 68% 72%
Ethnic Minority1 72% 51%
Married 26% 32%
Never Married 68% 63%
HIV Prevalence 16% 46%
Previous HIV test 7% 34%
1. Defined as a respondent not of Han ethnicity in China and not of Kinh ethnicity in Vietnam.
Table 2 Distance from border and ethnic minority percentage among injecting drug users in border area
China Ethnic minority P Value
Yes No
Distance from the Border
0 km 33 (83%) 7 (18%) P = .1118
19 km 15 (83%) 3 (17%)
57 km 162 (69%) 73 (31%)
Table 3 Distance from border and ethnic minority percentage among injecting drug users
Vietnam Ethnic minority P Value
Yes No
Distance from the Border
1 km 29 (73%) 11 (28%) P = .0017
3 km 20 (57%) 15 (43%)
12 km 29 (62%) 11 (38%)
13 km 29 (64%) 16 (36%)
14 km 86 (43%) 113 (57%)
Figure 2 shows the relationship between mean number of years injecting drugs and the site's distance from the border. There is a general gradient of decreasing mean years of drug injecting running from the site farthest from the border on the Vietnamese side (Lang Son City) to the site farthest from the border on the Chinese side (Ning Ming City). This gradient was statistically significant, p = -0.002. The estimated slope was -0.02, so the average number of years of injection use decreased by about a week (0.02 of a year) for each kilometer in the Chinese direction. This was based on a weighted linear regression model fitted to the mean number of years used in a site and distance to the border of the site. The weights were based on the sample size at each site.
Figure 2 Years of injection use against distance to the border.
Figure 3 shows the HIV prevalence by distance to the border for the sites in Vietnam and China for the baseline data. There is a strong gradient of decreasing prevalence from the largest Vietnamese site and the one farthest from the border (Lang Son City) to the largest Chinese site and the one farthest from the border (Ning Ming City). Only one site- (Tan Thanh in Vietnam) does not fall along this gradient. Overall baseline seroprevalence among IDUs was 46% in Lang Son and in 17% Ning Ming. This gradient was statistically significant, p < 0.001. The estimated odds ratio was 0.97, so the odds of an individual being HIV positive decreased by 2–3% for each kilometer in the Chinese direction. This relationship between the proportion of subjects in a site testing positive and distance to the border was modeled using logistic regression.
Figure 3 HIV prevalence against distance to border. Gradient plots at baseline.
Table 2 shows factors related to crossing the border (in the 6 months prior to the interview) among the subjects in Vietnam and China. Overall, 10.6% of subjects in China and 24.5% of subjects in Vietnam reported having crossed the border in the six months prior to the interview. Being a member of an ethnic minority group and living closer to the border were strongly related to crossing the border in both countries. In Vietnam, being HIV seropositive was negatively related to crossing the border, while in China being HIV seropositive was positively related to crossing the border (although this relationship did not reach statistical significance, p = .12). Being a member of an ethnic minority group was related to being HIV seropostive in China (19% HIV seropositive among ethnic minority subjects versus 10% seropositive among ethnic majority subjects, (p < 0.05). Ethnicity was not related to HIV status in among the Vietnamese subjects.
Discussion
In this report, we present data on injecting drug use and HIV infection among IDUs in Lang Son province, Vietnam and Ning Ming County, Guangxi Province, China from a baseline survey conducted before implementation of a peer-based cross-border HIV prevention intervention.
Several limitations should be noted. First, we were working with cross-sectional data, where longitudinal data from the initial spread of injecting drug use in the area would have certainly been preferable. Second, there are measurement and sample size limitations. It would have been helpful to have a measure of ease of travel to the border rather than simple physical distance to the border. Also it would have been helpful to have sufficiently large sample sizes so that the relationship between being a member of an ethnic minority group and being HIV seropositive could be examined within individual geographic sites. Despite these limitations, there are very clear patterns in the data.
There are similar gradients for mean length of injecting history and baseline HIV prevalence running in descending order from the Vietnamese site farthest from the border to the Chinese site farthest from the border. These patterns are consistent with the theory that both the practice of drug injection and the prevalence of HIV infection among IDUs spread from Northern Vietnam to Southern China along a major heroin trans-shipment route [5]. The patterns in our data suggest that, in some circumstances, it may be possible to reconstruct histories of the diffusion of injecting drug use and HIV among IDUs using cross-sectional data.
There is clearly a potential for further cross-border transmission of HIV in both directions. Our discussions with the peer educators in the cross-border HIV prevention project suggest three primary reasons for these IDUs crossing the border: 1. Obtaining higher quality/lower priced drugs, 2. Avoiding police pressure on drug injectors, which can be unpredictably variable and involves large-scale periodic crackdowns, and 3. Personal factors, such as migratory employment, commerce, and family ties. It would appear to be very difficult to reduce these reasons for IDUs crossing the border.
Because of the possibility of arrest, IDUs who cross the border are unlikely to carry needles and syringes with them, even if they are crossing the border at places without any supervision or inspection.
Prevention of risky injections among border crossing IDUs will require very good supplies of sterile injection equipment on both sides. If IDUs who crosses the border cannot readily access sterile injection equipment on both sides of the border, then their fellow IDUs will need to have sufficient supplies of sterile injection equipment for use by themselves and the IDUs who cross the border. This will require large-scale safer injection programs on both sides of the border.
The majority of subjects in this study belong to ethnic minority groups, primarily Zhuang in China and Tay and Nung in Vietnam. Ethnic minority IDUs were also overrepresented among the border crossers. The issues of ethnic minority status, injecting drug use, and HIV infection deserve much more research and policy development. Ethnic minority IDUs are more likely to be infected with HIV in many places, from African-American and Latino/a IDUs in New York City [10] to Roma in Eastern Europe [11] to First Nations in Vancouver [12] to Vietnamese in Australia [13] to Manipuris in India [14]. As noted above, there was a strong relationship between ethnic minority status and HIV serpositivity in Ning Ming (OR = 5.08 (95% CI 1.41,18.26, p = .013) [15]. Social stigmatization of ethnic minority communities may make them more vulnerable to illicit drug use, including injecting drug use. Employment discrimination against ethnic minority communities may increase the extent to which drug distribution occurs in these communities, and to which drugs are transported by minority community members. Persons belonging to ethnic minority groups also may have important factors facilitating international travel, such as social support systems and persons that speak the same language on the other side of international borders.
Conclusion
The data presented here illustrate many of the factors in the international diffusion of HIV among IDUs at modest geographic scale. (There is a total distance of 71 kilometers between the two most distant sites in the study). These include gradients of length of injecting drug use and HIV seroprevalence across the international border, border crossing by IDUs and its association with HIV infection for those crossing from China into Vietnam, overrepresentation of ethnic minority persons among the border crossers. Both injecting drug use and HIV among IDUs are already well established among IDUs on the Vietnamese side of the border and injecting drug use is well established on the Chinese side of the border. HIV is present among IDUs on the Chinese side of the border, but at lower seroprevalence levels than in Lang Son.
There are multiple reasons that people cross the border in this area, and it would not appear to be possible to stop IDUs from crossing the border or from injecting drugs across the border. Thus, HIV prevention goals must include increasing the safety of injections among border crossers (as well as reducing risk behavior among the IDUs who do not cross the border). This will require coordinated HIV prevention that increases the likelihood that IDUs will inject safely on both sides of the border. Such a program has been implemented in Lang Son and Guangxi provinces. It includes peer outreach, increased access to sterile injection equipment through syringe distribution and exchange and a pharmacy voucher program. IDUs may exchange used injection equipment for new needles/syringes or for vouchers that can be redeemed at local pharmacies for needles/syringes, sterile water, and condoms. IDUs may also directly receive new needles/syringes or pharmacy vouchers even if they do not return used equipment. The program also includes large-scale collection and safe disposal of used needles/syringes, general community education about drugs and HIV, and social support for people living with HIV or AIDS [7].
Reducing the international transmission of HIV among injecting drug users will require programs at the global, regional, national, and "local cross-border" levels. The local cross border programs will need to be coordinated on both sides of the border and on a sufficient scale that IDUs who cross the border will be able to readily obtain clean injection equipment on the other side of the border. The cross-border HIV prevention project currently being implemented in Lang Son Province and Ning Ming County, Guangxi offers an example of how such a coordinated approach can be implemented. Evaluation data being collected in Lang Son and Ning Ming will be used to gauge the effectiveness of the interventions.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
TH conceived of the study, the study design and coordination and assisted in the drafting of the manuscript. DCD participated in the design of the study and drafted and edited the manuscript. TH and DCD supervised the data analysis.
PF, PJ and RK performed the statistical analysis and participated in its design and coordination and participated in the writing and review of drafts of the manuscript.
WL, YC & DM supervised the implementation of the project and data collection and processing for the Chinese sites, and participated in the writing and review of drafts of the manuscript.
DN, TVH, LKV, NDT & KTB supervised the implementation of the project and data collection and processing for the Vietnamese sites, and participated in the writing and review of drafts of the manuscript.
All authors read and approved the final manuscript.
Table 4 Factors associated with crossing the border among injecting drug users in the China-Vietnam border area
Vietnam Crossed borders P Value
Yes No
Ethnic Minority
Yes 56 (31%) 125 (69%) P = .0015
No 27 (16%) 138 (84%)
HIV
Positive 20 (13%) 139 (87%) P <.0001
Negative 63 (34%) 124 (66%)
Distance from the Border
1 km 39 (98%) 1 (3%) P <.0001
3 km 3 (9%) 32 (91%)
12 km 8 (29%) 20 (71%)
13 km 12 (27%) 33 (73%)
14 km 21 (11%) 177 (89%)
China
Ethnic Minority
Yes 28 (13%) 181 (87%) P = .0053
No 2 (2%) 81 (98%)
HIV
Positive 8 (17%) 40 (83%) P = .1242
Negative 22 (9%) 219 (91%)
Distance from the Border
0 km 20 (51%) 19 (49%) P <.0001
19 km 3 (17%) 15 (83%)
57 km 7 (3%) 228 (97%)
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors gratefully acknowledge all of the health department and clinic staff, other public officials, peer educators, and pharmacists in Ning Ming County, Lang Son Province and Guangxi Province who are participating in and supporting this project. We would also like to acknowledge the support of the National Institutes on Drug Abuse, U.S. National Institutes of Health. Grant number 1R01DA1470301.
==== Refs
Stimson G Aceijas C Hickman M Dehne K Rhodes T Global Estimate of Injecting Drug Use: April 20-24 200; Melbourne, Australia. 2004
D'Aquila RT Williams AB Petersen LR Williams AE HIV seroprevalence among Connecticut intravenous drug users in 1986-87: Race/ethnicity as a risk factor for HIV seropositivity: ; Washington, DC. 1987 170
Weiss SH Ginzburg HM Goeddert JJ Biggar RJ Mohica BA Blattner WA Risk for HTLV-III exposure and AIDS among parenteral drug abusers in New Jersey.: ; Atlanta, GA. 1985
Stimson GV Reconstruction of sub-regional diffusion of HIV infection among injecting drug users in South-East Asia: Implications for prevention. AIDS 1994 8 1630 1632 7848607
Beyrer C Razak MH Lisam K Chen J Lui W Overland heroin trafficking routes and HIV-1 spread in South and Southeast Asia AIDS 2000 17 75 83 10714570 10.1097/00002030-200001070-00009
Hoang TV Son NT Trung NT al. A rapid assessment of HIV/AIDS situation and vulnerabilities related to drug use in Lang Son 2001 Hanoi, National AIDS Standing Bureau
Hammett TM Des Jarlais DC Liu W Ngu D Tung ND Vu Hoang T Van LK Donghua M Development and implementation of a cross-border HIV prevention intervention for injection drug users in Ning Ming County (Guangxi Province, China and Lang Son Province, Vietnam International Journal of Drug Policy 2004 14 389 398 10.1016/j.drugpo.2003.07.003
Stimson GV Des Jarlais DC Ball A Drug Injecting and HIV Infection: Global Dimensions and Local Responses 1998 London, UCL Press
Holland J Ramazanoglu C Scott S Sharpe S Thomson R Risk, power and the possibility of pleasure: young women and safer sex AIDS Care 1992 4 273 283 1525200
Des Jarlais DC Perlis T Arasteh K Hagan H Milliken J Braine N Yancovitz S Mildvan D Perlman D Maslow C Friedman SR "Informed altruism" and "partner restriction" in the reduction of HIV infection in injecting drug users entering detoxification treatment in New York City, 1990-2001 JAIDS 2004 35 158 166 14722449
Grund JP Verbraeck H Ofner P "Marel O Del, Kas Kamel, Le Romes Durvar" (God Hits Whom He Chooses; The Roma Gets hit Twice). An exploration of drug use and HIV risks among the Roma of Central and Eastern Europe: April 1- 5; New Delhi, India. 2001
Craib KJP Spittal PM Wood E Laliberte N Hogg RS Li K Heath K Tyndall MW O'Shaughnessy MV Schechter MT Risk factors for elevated HIV incidence among Aboriginal injection drug users in Vancouver Canadian Medical Association Journal 2003 168 19 24 12515780
Elliott JH Mijch AM Street AC Crofts N HIV, ethnicity and travel: HIV infection in Vietnamese Au with injection drug use J Clin Virol 2003 26 133 142 12600645 10.1016/S1386-6532(02)00112-9
Organization NAIDSC Combating HIV/AIDS in India 2000-2001 2001 New Delhi, Government of India
Hammet TM Johnston P Kling R Liu W Ngu D Tung ND Binh KT Dong HV Hoang TV Van LK Donghua M Chen Y Des Jarlais DC Correlates of HIV status among injection drug users in a border region of Southern China and Northern Vietnam J Acquir Immun Defic Syndr 2005 38 228 235 submitted 10.1097/00126334-200502010-00016
|
16120225
|
PMC1232855
|
CC BY
|
2021-01-04 16:28:58
|
no
|
BMC Public Health. 2005 Aug 24; 5:89
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-89
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-931613892410.1186/1471-2458-5-93Research ArticleMethodological aspects of a GIS-based environmental health inspection program used in the Athens 2004 Olympic and Para Olympic Games Hadjichristodoulou Christos [email protected] Elpidoforos S [email protected] Virginia [email protected] Matthew E [email protected] Efstathios [email protected] Georgios [email protected] Varvara [email protected] Jeni [email protected] Department of Hygiene and Epidemiology, Faculty of Medicine, University of Thessaly, Larissa, Greece2 National School of Public Health – Olympic Planning Unit (OPU), Athens, Greece2005 2 9 2005 5 93 93 11 3 2005 2 9 2005 Copyright © 2005 Hadjichristodoulou et al; licensee BioMed Central Ltd.2005Hadjichristodoulou 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 use of geographical information system (GIS) technologies in public health surveillance is gradually gaining momentum around the world and many applications have already been reported in the literature. In this study, GIS technology was used to help county departments of Public Health to implement environmental health surveillance for the Athens 2004 Olympic and Para Olympic Games.
Methods
In order to assess the workload in each Olympic county, 19 registry forms and 17 standardized inspection forms were developed to register and inspect environmental health items requiring inspection (Hotels, restaurants, swimming pools, water supply system etc), respectively. Furthermore, related databases were created using Epi Info 2002 and a geographical information system (GIS) were used to implement an integrated Environmental Health inspection program. The project was conducted in Athens by the Olympic Planning Unit (OPU) of the National School of Public Health, in close cooperation with the Ministry of Health and Social Solidarity and the corresponding departments of Public Health in all municipalities that were scheduled to host events during the Athens 2004 Olympic and Para Olympic games.
Results
A total of 44,741 premises of environmental health interest were geocoded into GIS databases and several electronic maps were developed. Using such maps in association with specific criteria, we first identified the maximum workload required to execute environmental health inspections in all premises within the eleven Olympic County Departments of Public Health. Six different scenarios were created for each county, based on devised algorithms in order to design the most effective and realistic inspection program using the available inspectors from each municipality. Furthermore, GIS applications were used to organize the daily inspection program for the Olympic games, provide coloured displays of the inspection results and link those results with the public health surveillance of specific cases or outbreak investigation.
Conclusion
Our computerised program exhibited significant efficiency in facilitating the prudent use of public health resources in implementing environmental health inspections in densely populated urban areas as well as in rural counties. Furthermore, the application of simple algorithms in integrating human and other resources provided tailored and cost-effective applications to different public health agencies.
==== Body
Background
The Athens 2004 summer Olympic and Para Olympic games, was the largest sport event ever held in Greece, and involved a large number of athletes and spectators gathered in the greater metropolitan area of Athens and four other Olympic Cities. Protecting the health of the athletes, officials, spectators and the population of Athens during the Olympic games, was one of the top priorities of the public health professionals and the Greek government itself. The Ministry of Health and Social Solidarity, as early as three years prior to the games had assigned the different tasks for public health preparation to the appropriate agencies. The National School of Public Health in Athens, in cooperation with the Greek Ministry of Health, put together a team of public health professionals in order to form the Olympic Planning Unit (OPU), which took the responsibility of coordinating the Environmental Health Inspection programs in the Olympic cities.
The idea of implementing a comprehensive environmental health inspection program for the Athens 2004 Olympic Games, using a geographical information system (GIS), was based on previous reports by epidemiologists and environmentalists regarding the spatio-temporal associations between environmental exposures and distribution of diseases in the population [1-4]. The importance of place in relation to population health has long been implicated in ancient as well as in modern epidemiology [5]. Using GIS software along with geocoding, researchers utilize powerful tools in exploring the relationship between geographic location and health [6]. Many environmental health studies examined the use of GIS in the field of geographical epidemiology by drawing up disease maps and conducting ecological analyses [7]. GIS technology has also been increasingly used in the mapping of exposures of certain environmental hazards, which might disproportionately affect human populations [8]. Maantay has described a GIS-based environmental equity study, conducted in the past decade, reviewing the spatial relationship between environmental pollution and health [9]. In addition, the value of GIS has been assessed for health care planning within research and practical applications [10], for planning of joint health and social care services [11], and for assessing areas with shortage of physicians [12].
Furthermore, specific capabilities of the GIS technology, which allow users to produce clear and accessible maps in association with geocoded data reports, constitute a powerful tool that were deemed extremely helpful to our main objective [13,14]. While GIS technology had also been used in the surveillance of communicable diseases (real time outbreak and disease surveillance – RODS), there was no previous report on the use of GIS in Environmental Health Inspection programs in previous Olympic Games [15].
In this report, we present the methodology used to develop a comprehensive environmental health inspection program for the Athens 2004 Olympic and Para Olympic Games, using GIS. In addition, we demonstrate several applications of the GIS databases in scheduling the environmental health inspections and allocating human resources according to the workload of specific geographic regions.
Methods
The Olympic planning unit of the National School of Public Health initiated the public health preparations in the field of environmental health by identifying a number of public health risks in order to design and coordinate an effective environmental health inspection program to protect the health of both visitors and residents during the Olympic period. Food safety, drinking, and recreational water safety, as well as other items such as pest control, waste management (solid and liquid), legionella prevention (water supply systems, cooling towers, decorative fountains), public toilet sanitation and vessel sanitation were determined as the main items of public health interest and the most appropriate targets of the environmental health inspection program. According to experience from previous Olympic Games and other mass gatherings, the most frequent public health problems included food borne and/or waterborne outbreaks [15,16].
In order to design and coordinate an environmental health inspection program, the National School of Public Health – Olympic Planning Unit (NSPH – OPU), developed a close cooperation with a network of public health professionals from the office of environmental health at the Ministry of Health, the office of medical services from the Athens 2004 Olympic Games Organizing Committee and all county departments of public health responsible to oversee inspection programs in municipalities hosting Olympic events.
Olympic county departments of public health
The vast majority of Olympic events were scheduled to take place in the metropolitan area of Athens, which included four sections of the municipality of Athens (central, east, west and south), the municipalities of East and West Attica, and the municipality of Piraeus with its main port. Additional counties hosting Olympic events included the metropolitan area of Thessalonica with its second largest port in the north of the country (the second largest city of Greece), the metropolitan area of Patra with its port, and two smaller cities; the city of Volos and the city of Iraklio. Environmental health inspectors from the seven County Departments of Public Health (CDPH) of the greater metropolitan area of Athens and the CDPH of the other four Olympic cities (Thessalonica, Patra, Volos, and Iraklio) were responsible for conducting the environmental health inspections prior to, and during the Olympic and Para Olympic Games.
Workload assessment
To assess the workload of county departments of public health (CDPH) located in Olympic cities, 19 specific registry forms and 17 standardized inspection forms were created. Public health inspectors identified and registered all premises of public health interest in each Olympic CDPH, in advance, via on-site visits, using the registry forms. The premises targeted were the following: Olympic Venues, hotels, archaeological sites, cruise ships, camps, seacoasts, airports, marines, bottle-water plants, ice-producing plants, restaurants, other food premises, canteens, water supply systems for buildings and vessels, swimming pools, decorative fountains, cooling towers, waste management facilities, sewage treatment units, public toilets, and areas requiring pest control. Corresponding to the above-described forms, related databases were created using Epi Info 2002 in order to record basic information for the eligible inspection items and document the inspection results [17].
The exact address and postal code of each premise of environmental health interest and each archaeological and tourist area were used to develop GIS maps for each county. By employing the Arc view 3.2 software, all registered addresses were geocoded into specific databases [18]. The longitude and latitude coordinates for every geocoded address were assigned automatically from the digital map (called a street reference map), so that we would visualize the location of each specific place. Using the above maps, we estimated the total miles along with the mean inspection time required to inspect each registered premise, by taking into account the inspection time itself and the one-way transportation time to each premise. The transportation time was calculated using GIS simulated travel time on a particular street network. Pedestrian streets, one-way roads and specifically designated Olympic lanes and other restrictions employed during the Olympic games were taken into consideration for the calculation of the transportation time.
The following criteria were considered as constant parameters used to estimate the workload of each county (1) all inspections are performed by two inspectors (a pair). The assignment of inspectors in pairs was based on their experience in inspecting food premises or water sites in order to achieve complementary expertise; (2) the daily working time is estimated at 6 hours including transportation and inspection time; (3) the inspectors' transportation speed was estimated, using our own inspection data. Driving was estimated at about 10 kilometers (km) per hour (h), and walking on a pedestrian zone at about 4 km/h, and on stairways at about 2 km/h, respectively; and (4) the starting point of every daily route was considered to be each CDPH office. Using the above criteria, we calculated the total number of inspection hours needed for each CDPH. The number of working days, for every inspector, was estimated by dividing the total number of estimated inspection hours by six hours, which represented the daily working time for inspectors. In addition, we estimated the required person-time (in months) for the total number of inspections (each month was considered as 22 working days).
Designing the environmental health inspection program
The Olympic environmental health inspection program had two main components for two time periods (the Pre-Olympic and the Olympic period). The first component referred to the inspections inside the Olympic Venues and the second component to the inspections outside the Olympic Venues. Therefore, the program was modified for each component and time period accordingly. Environmental health inspections inside the Olympic Venues were scheduled to be conducted strictly by certified inspectors. All environmental health inspections included on-site inspections and selective sampling for chemical and microbiological examinations. In order to achieve high quality and operational stability, the inspections were standardized and were performed based on standardized checklists and sampling forms. Seventeen standardized inspection forms were created using a negative scoring system [see additional files 1, 2]. All inspected premises were ranked in three categories (A, B or C) according to their score. Category A represented the highest score and therefore the best premises with respect to public health risks.
In order to cope with the enormous number of existing food premises (restaurants, bars, canteens) in the area of Athens, and devise the most realistic and cost-effective environmental health inspection program for the Olympic Games, a series of scenarios was developed, including a different number of food premises in each scenario, which depended on the available number of inspectors, cars and financial resources of each Olympic CDPH.
The first scenario included inspections of all premises in all eleven Olympic CDPHs. In this first scenario, a large number of inspectors would be required to work for a long period of time in order to cover the total number of inspections. For example, in the central section of Athens, inspecting all 8,358 premises would require five pairs of inspectors working for 24.9 months. Apparently this scenario was not realistic because it would impede counterchecks of several locations whenever needed. The second scenario estimated the time and personnel required for the inspection of hotels and Olympic Venues only. However, this would not be sufficient for protecting public health during the Olympic and Para Olympic Games. The third, fourth and fifth scenarios included the inspection of all premises in each Olympic county and the inspection of all food premises around Olympic Venues and Olympic hotels in buffers of 1,000, 500 and 300 meters, respectively. The above scenarios enabled us to conduct a sufficient number of inspections; however a big number of food premises were not included. Finally, the sixth scenario, which was recommended and selected for implementation, included the following: (1) inspection of all premises with the exception of food premises, in each Olympic CDPH; (2) inspection of all food premises around Olympic Venues and Olympic hotels in a buffer of 200 meters; (3) inspection of all food premises around touristy and archaeological areas in a buffer of 200 meters; and (4) inspection of 2% of the food premises, randomly selected, from the rest of each county.
Partitioning of the Olympic counties
An algorithm was developed and utilized based on GIS technology in order to partition each Olympic CDPH in as many equal parts as the number of available pairs of inspectors. The term equal did not refer to the geographical area or the number of premises eligible for inspection. Rather the partitioning was based on the workload of each CDPH (required time for inspecting all identified locations of environmental health interest). Therefore, we organized the daily inspection program in such a way that the available inspectors in each county would perform the maximal possible number of inspections during the Pre-Olympic and Olympic period.
Other GIS applications of the program
By using databases from Epi info 2002 and its GIS component, we were able to develop additional applications in order to implement the daily inspection program. For example, we were able to display the inspection results in color codes according to the categorical score received by each premise. Furthermore, we had the ability to link the inspection results to information regarding the surveillance of human cases of Legionnaire's disease.
Results
Olympic counties' workload
A total of 44,741 premises of environmental health interest were registered using several reporting forms. As a result, 44,741 addresses were geocoded into GIS databases and developed into appropriate working maps. In Table 1 we present the total number of registered premises in each Olympic County Department of Public Health. In Figure 1 we provide an example of a GIS map produced, representing all premises of environmental health interest of the central section of Athens, which was the most important Olympic county since most of the Olympic events took place within its boundaries.
Table 1 Total number of registered premises of environmental health interest in each Olympic County Department of Public Health (CDPH) in Greece
County Department of Public Health Olympic Venues Hotels Food Premises Other Premises Total
Central Athens 5 254 7,954 145 8,358
Eastern Athens 5 22 2,117 133 2,277
Southern Athens 3 53 2,467 125 2,648
Western Athens 1 5 2,570 108 2,684
Eastern Attica 6 80 1,961 100 2,147
Western Attica 1 15 1,463 57 1,536
Piraeus 3 258 5,356 224 5,841
Thessalonica 1 111 9,927 57 10,096
Patra 1 100 3,112 49 3,262
Volos 1 238 2,522 77 2,838
Iraklio 1 466 2,484 103 3,054
Total 28 1602 41,933 1,178 44,741
Figure 1 The total number of registered premises of environmental health interest represented by red dots (Olympic Venues, hotels, restaurants, camps, swimming pools, cooling towers etc.) in the central section of Athens, which were eligible for inspection.
Olympic inspection program
Based on the sixth scenario described in the methods section, we determined the exact number and type of premises that were to be inspected in each CDPH during the Pre-Olympic and Olympic period. For instance, the workload of the central section of Athens was estimated at 1,250 hours or 420 person-days for the inspection of 1,010 environmental health items. Therefore, the total inspection time for 5 pairs of inspectors was estimated at 4.7 months. In Table 2 we compare the application of two different scenarios; the first and the sixth scenario; for all Olympic CDPH and in Figure 2 we present a GIS image of the recommended scenario in the central section of Athens.
Table 2 Comparison of scenario 1 (inspection of all county premises) to scenario 6 (inspection of a selected number of premises) in each Olympic CDPH in Greece
Scenario 1 Scenario 6
County Department of Public Health Number of Premises for inspection Required time (in months) Number of Premises for inspection Required time (in months)
Central Athens 8358 24.9 1010 4.7
East Athens 2277 9.1 578 2.8
South Athens 2648 9.4 929 3.8
West Athens 2684 9.6 640 2.9
East Attica 2147 8.8 386 2.6
West Attica 1536 5.7 171 2.4
Piraeus 5841 17.9 1532 6
Thessalonica 10096 29.7 599 3.5
Patra 3262 11.4 344 3.6
Volos 2838 10.4 429 4.3
Iraklio 3054 11.9 704 4.9
Total 44741 148.8 7322 41.5
Figure 2 The total number of premises represented by small red dots (Olympic Venues, hotels, restaurants, camps, swimming pools, cooling towers etc.) requiring inspection in the central section of Athens (shaded in grey), as defined by the computerized program applying the sixth scenario (described in the text). Buffers with increasing size are displaying areas of public health interest around Olympic Venues and tourist and archaeological sites.
Partitioning
The partitioning of every Olympic county was based on the available number of inspectors (pairs) employed in each county. For example, the central section of Athens, where five pairs of inspectors were employed, was partitioned in five parts with equal workload as it is shown in Figure 3.
Figure 3 Partitioning of the central section of Athens in five parts with equal workload in terms of premises of environmental health interest requiring inspection based on the sixth scenario. Each part is displayed by a different color representing the workload of each pair of inspectors from the specific county.
Other GIS applications
The results of all environmental health inspections and all laboratory tests (microbiological examinations of the field samples obtained), were displayed on the electronic maps, in different colors according to the inspection score (score categories A, B or C), so that we would be able to automatically detect those premises that failed inspection and schedule counterchecks. For example, as we present in Figure 4 for the whole country and in figure 5 for the central section of Athens, 272 out of 781 premises of environmental health interest (34.83%) in the central section of Athens, inspected for the first time, received score A, 270 (34.57%) of the premises received score B and the remaining 239 (30.60%) received score C. All premises receiving score C had to undergo counterchecks within a month from the first inspection.
Figure 4 (A) Screenshot of Epi Map (part of US CDC Epi Info) [17] showing inspection results, on a zoomable GIS map, of all Olympic cities in Greece. The dark red dots and dark pink colored dots represent the premises, which received inspection score A and B, respectively (satisfactory results). The light pink colored dots represent the premises with the worst inspection score (score C – unsatisfactory results requiring counterchecks). (B) Screenshot of Epi Map (part of US CDC Epi Info) [17] showing inspection results, on a zoomable GIS map, of the central section of Athens. The dark red dots and dark pink colored dots represent the premises, which received inspection score A and B, respectively (satisfactory results). The light pink colored dots represent the premises with the worst inspection score (score C – unsatisfactory results requiring counterchecks). White dots represent environmental health items that were not inspected. Information on the inspection scores and laboratory results were available to be viewed on the map by selecting a specific environmental item as indicated by the window on the bottom left of the figure.
Figure 5 (A) Screenshot of Epi Map (part of US CDC Epi Info) [17] showing inspection results, on a zoomable GIS map, of the central section of Athens. The dark red dots and dark pink colored dots represent the premises, which received inspection score A and B, respectively (satisfactory results). The light pink colored dots represent the premises with the worst inspection score (score C – unsatisfactory results requiring counterchecks). White dots represent environmental health items that were not inspected. Information on the inspection scores and laboratory results were available to be viewed on the map by selecting a specific environmental item as indicated by the window on the bottom left of the figure.
Figure 6 Screenshot of Epi Map (part of US CDC Epi Info) [17] showing all premises of environmental health interest (cooling towers, decorative fountains etc) in a buffer of 200 meters around a hypothetical case of legionnaires' disease in the central section of Athens depicted by a blue square (patients' residence). The dark red dots and dark pink colored dots represent the premises, which received inspection score A and B, respectively (satisfactory results). The light pink colored dots represent the premises with the worst inspection score (score C – unsatisfactory results). GIS map features included the ability to review, in real time, the inspection scores and laboratory results by selecting specific items within the buffer zone.
Another application was the correlation of potential cases of Legionnaires' disease in humans, with specific cooling towers or the water supply system of a specific building or even a decorative fountain in a specific region. For example, as it is presented in Figure 6, if a case of Legionnaires' disease would be detected in a person, who lived in a given address, all buildings of public health interest around the specific location, and included in a buffer of 200 meters would be evaluated automatically for previous inspection scores and laboratory results for legionella. Using this method the most possible source of infection could be identified.
Discussion
To our knowledge, this is the first application of GIS technology that was used to plan, organize and implement the environmental health inspection program for the Olympic Games. The application of GIS technology along with data management and analysis, using the Epi info 2002, in the implementation of the Olympic environmental health inspection program, has been a valuable tool in selecting a realistic and cost-effective program that covered the most important premises of every Olympic county department of public health. Since visitors, spectators and athletes were expected to move, mostly, around hotels, Olympic Venues and highly tourist places, we believe that the recommended GIS-based inspection program represented one of the best approaches aiming to satisfy the demands of the Olympic Games in Athens while focusing on areas of sporting and entertaining events.
Moreover, a relatively low percentage (2%) of food premises outside the target geographical regions (buffers around Olympic venues and archaeological sites), were randomly selected for inspection in the pre-Olympic as well as during the Olympic period. In an effort to improve the beneficial effect of the inspections and diffuse the word around the Olympic cities about the environmental health inspection program, the results of the pre-Olympic period were publicized by the National Food Authority of the Ministry of Development.
Several useful applications of the GIS technology in Greece have been previously described such as in the development of operational systems to support decisions during large forest fire incidents [19]. In addition, it has been used for the implementation of an active surveillance program for brucellosis in a rural area of central Greece [20] and in the description of the profile of pollution of drinking water by nitrates, chloride and arsenic in Northern Greece [21]. We believe that the current application for the environmental health inspection program for the Athens 2004 Olympic and Para Olympic Games constituted a novel idea, which may be followed in future sporting events around the world.
We would like to acknowledge several limitations of our study. One possible source of error is the estimation of the workload of each county, which is based on criteria such as the inspectors' transportation speed, their daily working hours and the required inspection time for each item. For example, we are unable to predict, with reasonable accuracy, the speed of the inspectors and the duration of the inspections because both can be affected by unpredictable factors. In addition, two hours out of the six working hours of the inspectors were needed for completing the official reports used for each inspection and for other related subjects. Furthermore, the partition of the Olympic counties could be implemented taking into consideration primarily the distance and secondly the workload; however in that case, it would be very difficult to create a small number of partitions. Ideally, the results of the environmental health inspection program with GIS technology should be linked to the communicable disease surveillance for human cases. However, the above linkage was operational only for possible cases of legionella. A fully operational system for all communicable diseases was not in place due to limitations of human surveillance data (lack of geocoding).
Conclusion
Despite the above restrictions, our program has proven its value whenever was tested during different events in the pre-Olympic and the Olympic period. Using our GIS-based inspection program, we have succeeded in covering all test events for the Olympic Games one year prior to the actual Olympic period with the least number of inspectors. Our program offered several advantages to public health authorities in Greece and may prove useful in designing, planning and implementing environmental health inspection programs in densely populated urban areas as well as in rural counties given its flexibility. In addition, the integration of human and other resources in simple algorithms provides tailored and cost-effective applications to each public health region. Furthermore, the GIS-based tools may support several other functions, described above, that allow inspectors to visualize the results on a colour scale, as well as associate them with other findings of public health importance such as data on outbreaks and or epidemics from public health surveillance. We believe that our method, as applied in the eleven Olympic counties in Greece, is relatively flexible and if accordingly modified, could be used by environmental health authorities in future Olympic Games and other mass gatherings around the world.
Abbreviations
OPU – Olympic Planning Unit
GIS – Geographical Information System
CDPH – County department of public health
Km/h – Kilometers per hour
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CH, VK, EP, GP, and VM carried out the data collection. CH, ESS and VK drafted the manuscript. MEF, JK reviewed the manuscript and made significant comments. CH, VK, EP, GP, VM and JK participated in the design of the study. CH, EP and GP performed the statistical analysis. CH and JK conceived of the study, and along with EP and GP participated in its coordination. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional file 1
Translated version of the standardized inspection form for restaurants
Click here for file
Additional file 2
Translated version of the standardized inspection form for mobile canteens
Click here for file
Acknowledgements
The program was funded by the Greek Ministry of Health and Social Solidarity.
==== Refs
Vine MF Degnan D Hanchette C Geographic information systems: their use in environmental epidemiologic research Environmental Health Perspectives 1997 105 598 605 9288494
Moore DA Carpenter TE Spatial analytical methods and geographic information systems: use in health research and epidemiology Epidemiologic Reviews 1999 21 143 161 10682254
Bellander T Berglind N Gustavsson P Jonson T Nyberg F Pershagen G Jarup L Using geographic information systems to assess individual historical exposure to air pollution from traffic and house heating in Stockholm Environmental Health Perspectives 2001 109 633 639 11445519
Miranda ML Dolinoy DC Overstreet MA Mapping for prevention: GIS models for directing childhood lead poisoning prevention programs Environmental Health Perspectives 2002 110 947 953 12204831
Bonner MR Han D Nie J Rogerson P Vena JE Freudenheim JL Positional accuracy of geocoded addresses in epidemiologic research Epidemiology 2003 14 408 412 12843763
McElroy JA Remington PL Trentham-Dietz A Robert SA Newcomb PA Geocoding Addresses from a Large Population-based Study: Lessons Learned Epidemiology 2003 14 399 407 12843762
Kistemann T Dangendorf F Schweikart J New perspectives on the use of Geographical Information Systems (GIS) in environmental health sciences International Journal of Hygiene and Environmental Health 2002 205 169 181 12040915 10.1078/1438-4639-00145
Maantay J Mapping environmental injustices: pitfalls and potential of geographic information systems in assessing environmental health and equity Environmental Health Perspectives 2002 110 161 171 11929725
Barnes S Peck A Mapping the future of health care: GIS applications in health care analysis Geographical Information Systems 1994 4 30
Foley R Assessing the applicability of GIS in a health and social care setting: planning services for informal carers in East Sussex, England Social Science and Medicine 2002 55 79 96 12137191 10.1016/S0277-9536(01)00208-8
Luo W Using a GIS-based floating catchment method to assess areas with shortage of physicians Health and Place 2004 10 1 11 14637284 10.1016/S1353-8292(02)00067-9
Boulos MN Towards evidence-based, GIS-driven national spatial health information infrastructure and surveillance services in the United Kingdom Int J Health Geogr 2004 28; 3 1 14748927 10.1186/1476-072X-3-1
Boulos MN Roudsari AV Carson ER Health geomatics: an enabling suite of technologies in health and healthcare J Biomed Inform 2001 34 195 219 11723701 10.1006/jbin.2001.1015
Gesteland PH Gardner RM Tsui FC Espino JU Rolfs RT James BC Chapman WW Moore AW Wagner MM Automated syndromic surveillance for the 2002 Winter Olympics J Am Med Inform Assoc 2003 10 547 54 12925547 10.1197/jamia.M1352
Banwell K Environmental health preparation for the Sydney 2000 Olympic and Para Olympic Games NSW Public Health Bulletin 2000 11 147 148 12105460
Hadjichristodoulou C Mouchtouri V Soteriades ES Vaitsi V Kolonia V Vasilogiannacopoulos AP Kremastinou J Mass gathering preparedness: the experience of the Athens 2004 Olympic and Para-Olympic Games J Environ Health 2005 67 52 57 15957323
Epi Info
Arc View 3.2
Keramitsoglou I Kiranoudis CT Sarimvels H Sifakis N A Multidisciplinary Decision Support System for Forest Fire Crisis Management DSS for forest fire crisis management Environmental Management 2004, Mar 23
Hadjichristodoulou C Soteriades E Goutzianna G Loukaidou M Babalis T Antoniou M Delagramaticas J Tselentis Y Surveillance of brucellosis in a rural area of Greece: application of the computerized mapping programme European Journal of Epidemiology 1999 15 277 283 10395059 10.1023/A:1007455923547
Fytianos K Christophoridis C Nitrate, arsenic and chloride pollution of drinking water in Northern Greece. Elaboration by applying GIS Environmental Monitoring and Assessment 2004 93 55 67 15074609 10.1023/B:EMAS.0000016791.73493.aa
|
16138924
|
PMC1232856
|
CC BY
|
2021-01-04 16:28:57
|
no
|
BMC Public Health. 2005 Sep 2; 5:93
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-93
|
oa_comm
|
==== Front
Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-501611782710.1186/1475-925X-4-50ResearchRemoval of power-line interference from the ECG: a review of the subtraction procedure Levkov Chavdar [email protected] Georgy [email protected] Ratcho [email protected] Ivan [email protected] Ivaylo [email protected] Ivan [email protected] Signa Cor laboratory, Lubliana str 46, 1618 Sofia, Bulgaria2 Technical University of Sofia, Faculty of Electronic Engineering and Technologies (FETT), Kliment Ohridski str. 8, 1000 Sofia, Bulgaria3 Centre of Biomedical Engineering, Bulgarian Academy of Sciences, Acad. G. Bonchev str., blok 105, 1113 Sofia, Bulgaria2005 23 8 2005 4 50 50 16 6 2005 23 8 2005 Copyright © 2005 Levkov et al; licensee BioMed Central Ltd.2005Levkov 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
Modern biomedical amplifiers have a very high common mode rejection ratio. Nevertheless, recordings are often contaminated by residual power-line interference. Traditional analogue and digital filters are known to suppress ECG components near to the power-line frequency. Different types of digital notch filters are widely used despite their inherent contradiction: tolerable signal distortion needs a narrow frequency band, which leads to ineffective filtering in cases of larger frequency deviation of the interference. Adaptive filtering introduces unacceptable transient response time, especially after steep and large QRS complexes. Other available techniques such as Fourier transform do not work in real time. The subtraction procedure is found to cope better with this problem.
Method
The subtraction procedure was developed some two decades ago, and almost totally eliminates power-line interference from the ECG signal. This procedure does not affect the signal frequency components around the interfering frequency. Digital filtering is applied on linear segments of the signal to remove the interference components. These interference components are stored and further subtracted from the signal wherever non-linear segments are encountered.
Results
Modifications of the subtraction procedure have been used in thousands of ECG instruments and computer-aided systems. Other work has extended this procedure to almost all possible cases of sampling rate and interference frequency variation. Improved structure of the on-line procedure has worked successfully regardless of the multiplicity between the sampling rate and the interference frequency. Such flexibility is due to the use of specific filter modules.
Conclusion
The subtraction procedure has largely proved advantageous over other methods for power-line interference cancellation in ECG signals.
==== Body
Background
Modern biomedical amplifiers have very high common mode rejection ratio (CMRR), with commercial ECG instruments manifesting values up to 120 dB. Nevertheless, recordings are often contaminated by residual power-line (PL) interference. This is due to differences in the electrode impedances and to stray currents through the patient and the cables. Thus, the common mode voltage is transformed into a false differential signal [1-4] that cannot be suppressed even by an infinitely high CMRR. The problems become more complicated if the instrument has a floating input to increase patient safety [5,6].
CMRR of a commercial ECG instrument is typically measured under laboratory conditions using generators with low impedance output and short connecting wires. Thus, a claim of CMRR > 60 ÷ 70 dB in the real world of ECG acquisition is without legitimate basis.
Any residual PL interference may interfere with the correct delineation of ECG wave boundaries [7] and corrupt the proper function of automatic ECG analysis. The interference can also disturb the correct measurement of RR intervals, which is the basis for heart rate variability analysis.
Hardware solutions have been developed to increase the actual CMRR by equalization of the cable shield and the right leg potentials [4]. This reduces the influence of stray currents through the body, but the efficiency obtained is not sufficient to significantly reduce the interference.
Traditional analogue and digital filters are known to suppress ECG components near the PL frequency. Different types of digital notch filters are widely used [8,9] despite their inherent contradiction: tolerable signal distortion needs a narrow frequency band, which leads to ineffective filtering with larger PL frequency deviation. Moreover, the resulting transient time is often unacceptably long. Hamilton [10] compares the convergence times of adaptive and non-adaptive notch filters. Both introduce significant distortion in the QRS and ST-segment portions due to the filter ringing. Soo-Chang and Chien-Cheng [8] try to reduce to some extent the transient response time by using vector projection to find better initial values for IIR notch filters. Yoo et al. [11] propose a hardware notch filter with adaptive central frequency to follow the PL frequency changes, thus defining a narrower bandwidth. Filters with various Q factors have been tried. However, the resulting signal distortion cannot be correctly assessed because of the reduced scale of the examples provided [11].
Instead, the condition was simulated in the MATLAB environment [12]. A synthesized ECG signal (without noise) was mixed with constant 1 mVp-p 50 Hz interference and processed by notch filters with bandwidths: 49–51, 49.5–50.5, and 49.9–50.1 Hz. This 1 Hz bandwidth is one order of magnitude narrower than that used by Yoo et al. [11]. Acceptable distortion was found only with the 49.9 ÷ 50.1 Hz filter, but after an exclusively long tail of about 12 s. This adaptation period reappeared with abrupt power-line frequency change of 0.2 Hz, despite a synchronised identical shift of the filter centre frequency.
Ringing is also present when spectral components of the interference are removed from the ECG signals using the Fourier transform [13]. Furthermore, this transform does not work in real time.
Mitov [14] uses parabolic detrending of ECG to estimate the signal components with frequencies corresponding to PL interference by using the discrete Fourier transform, to approximate averaged interference values, which are subtracted from the contaminated signal. No results with frequency-modulated interference are presented in the publication.
The interference may be suppressed by adaptive filtering [15-17]. However, this technique introduces unacceptable transient response time, especially following signals of steep and high amplitude, e.g. the QRS complex.
Kumaravel and Nithiyanandam [18] reported interference cancellation by an off-line working genetic algorithm.
Some authors do not present the results of their algorithms correctly or clearly enough to use for interference removal. Sometimes the original signal is not presented [19], no differences between original and processed signals are shown [20], and the performance is measured by the error square instead of amplitude differences [21].
Method
The subtraction procedure for PL elimination was first elaborated some two decades ago [22]. This procedure does not affect the ECG components neighbouring the PL frequency. This theoretical study is carried out for the basic PL frequency, but the conclusions are also valid for its harmonics and, consequently, for an arbitrary interference waveform. The efficiency of the procedure does not depend on the amplitude of the interference, as long as the amplifier is not saturated. Moreover, the procedure copes successfully with changes in amplitude and frequency of the interference. The procedure has been continuously improved over the years [12,13,23-32], and implemented in thousands ECG instruments and computer-aided systems [33,34]. Similar approaches have also been published by other researchers [35-41].
Principles
The subtraction procedure is applied originally with sampling frequency fS, a multiple of, and hardware synchronized with the PL frequency fPL. The procedure consists of the following steps [22]:
• ECG segments with frequency band near zero are continuously detected using an appropriate criterion. They are referred to as linear segments and are found mainly in the PQ and TP intervals, but also in sufficiently long straight parts of the R and T waves.
• The samples of these segments are moving averaged, i.e., subjected to a linear phase comb filter [42] with first zero set at fPL. Thus, the filtered samples do not contain interference.
• Interference amplitudes, called corrections, are calculated for each of the phase-locked samples, n, in the PL period, TPL, by subtracting the filtered samples from the corresponding ones of the contaminated (original) ECG signal.
• The set of corrections obtained is continually updated in linear segments and used in non-linear segments (usually around QRS complexes and high-amplitude T waves) to subtract the interference from the original ECG signal.
One of the first results obtained by the subtraction method is shown in Fig. 1 [22]. Interference was added to a clean simulated ECG signal in order to evaluate the errors and the efficiency of the method.
Figure 1 One of the first results obtained by the subtraction procedure.
Linear criterion
A linear criterion, Cr, usually corresponds to the second difference of the signal (mathematical evaluation of the linearity). The first Cr [22] is defined in the following manner. Six consecutive first differences, FDi, are calculated using signal samples, Xi, spaced at one TPL:
FDi = Xi+n - Xi, for i = 1 ... 6 (1)
The PL interference in the first differences is suppressed if n = fS/fPL. In this case n = 5, since the procedure was developed initially for rated fPL = 50 Hz and fS = 250 Hz. Furthermore, the maximum FDmax and minimum FDmin values are taken to determine Cr:
Cr = | FDmax - FDmin | <M, (2)
where M is the threshold value.
Typical linear and non-linear segments are shown in Fig. 2. Real ECG signal (trace a) is superimposed by interference (trace b). The linear segments include low frequency signal and power-line frequency components. An approximate frequency spectrum of such linear segments is shown in Fig. 3.
Figure 2 Typical linear and non-linear segments in real ECG signal.
Figure 3 Approximate frequency spectrum of a linear segment.
This criterion works accurately, but can hardly be applied in real time because its relatively slow implementation. This drawback is overcome by Christov and Dotsinsky [23] who use a modified criterion of just two subsequent differences.
Cr = | FDi+1 - FDi | <M. (3)
The first sample, which does not fulfill equation (3), is associated with the beginning of a non-liner segment. In the non-linear to linear transition, equation (3) should be satisfied consecutively n times in order to avoid premature detection of the linear segment. The criterion is implemented in real time for fS = 400 Hz and n = 8.
Later, Dotsinsky and Daskalov [13] defined the criterion as two non-subsequent differences:
Cr = |FDi+k - FDi | <M, for k >1 (4)
This approach makes the transition from linear to non-linear segment more precise.
Compensation of PL amplitude variations
The more frequently the corrections are updated, the better compensation of the amplitude variations of the PL is achieved. Therefore, the linear criterion threshold, M, has to be reasonably less restrictive so that the errors, committed by averaging some segments that depart from the ideal linear signal, are smaller than the errors, that will appear if M initiates sporadic updating of the correction. Initially, M was fixed at 160 μV [22]. Later, heuristic values were found to be M = 150 μV [23] and M = 100 μV [13].
Linear filtering
For odd sample number n = 2m + 1 in one period of the PL interference, the filtered value:
is phase-coincident with the non-filtered one.
In case of even number n = 2m, the two values are phase-shifted by a half of the sample period:
but become in-phase coincidence using the formula
It is possible to take for averaging every second, third or qth sample if n/q is integer. Depending on whether n/q is odd or even, equation (7) or (8) is used, respectively.
A special case of maximum sample reducing arises with q = n/2 [28]. The corresponding formula:
is called 'three-points' filter. In addition to equation (8), the following formula
can also be applied if q is even. In case of q = n/2, the filter becomes 'two-points' and is represented by:
Reduced sample number in a period of the interference will lead to enhanced steep slope of the comb filter lobes and will shorten the computation time. However, these 'advantages' must be assessed carefully in order not to violate the Nyquist rule with a large amount of the third harmonic present. The other harmonics are not taken into consideration since the highest odd harmonics are usually suppressed by low-pass filters with cut-off in the range of 100–150 Hz, while the even ones are practically absent because of the precise pole manufacturing of the electric power station generators.
Compensation of PL frequency variation
The allowed deviation from the rated PL frequency is limited in some countries up to 1% by the standards. In practice, deviation is oftentimes higher. Kumaravel et al. [43] reported for variation of 3%. McManus et al. [44] found considerable changes in the interference frequency, which is superimposed on recordings taken from the Common Standards for Electrocardiography (CSE) database.
Frequency variations lead to a special case of non-multiple sampling with real n, instead of integer one. This complication can be bypassed if the deviations are detected by continuous hardware measurement of fPL and corrected by small adjustments of the sample interval tS around its rated (R) value, tRS = TRPL/n (here, TRPL = 20 ms is the rated TPL for fRPL = 50 Hz). For fPL, deviation between 49.5 and 50.1 Hz, the tS variations are in the range of 1%, and consequently they do not introduce errors beyond the accepted measuring accuracy of parameters that are usually used for automatic ECG classification.
A first approach associates the triggering of each first sample, S1, of the sequences Sk (k = 1, 2...n) in the periods TPL with arbitrary chosen but constant amplitude of the PL voltage. The next samples, Sk (k = 2...n), are spaced at tS, which is obtained by tS = TRTL/n. For 50 Hz, and n = 5, tS = 4 ms. Two types of errors committed using this approach are studied by Dotsinsky and Daskalov [13]. The first, due to inter-sample irregularities, may reach 1% at fS = 400 Hz and 1.2% at fS = 250 Hz, in case of 1% deviation around the fRPL. The second type of error does not exceed 3% and is a consequence of the additionally shifted location of the filtered sample.
Dotsinsky and Daskalov [13] reported an improved approach. The ongoing period TPL is measured and divided by n. The obtained tS is used in the subsequent TPL.
Efficiency assessment of the procedure
Subtraction procedure examples are shown in Fig. 4 and Fig. 5. The ECG signals are taken from the American Heart Association database. The signals are notch filtered to suppress the 60 Hz (PL frequency in the US) interference. Then, the signals are mixed with 50 Hz interference, amplitude modulated from 0 through 3.2 mVp-p by a slew rate of 200 μVs-1. The traces are identified as follows: i) input signal; ii) synthesized interference; iii) mixed signal; iv) processed signal; v) difference between original and processed signals and vi) zoomed difference. Actually, the discarded components also include electromyographic (EMG) and other noises. A non-suppressed part of the signal, together with small residual interference and distortions due to compromise with the M value are also present in the differences.
Figure 4 Processed AHA 3004d1 recording.
Figure 5 Processed AHA 6007d1 recording.
Two signals are used to assess the efficiency of the subtraction procedure with respect to the interference only. The first, taken from our own database, is called conditionally 'clean' (Fig. 6). The result shows small differences between input and processed signals, visually due to the noise presented in the input. This result is verified with the second synthesized signal, which does not contain any disturbances (Fig. 7). As can be seen, no distortions are introduced. The same synthesized signal is superimposed by interference and processed (Fig. 8). No residual interference can be found.
Figure 6 Processing of conditionally 'clean' signal.
Figure 7 Processing of synthesized signal.
Figure 8 Processed synthesized signal+interference.
Influence of EMG noise
Sometimes EMG noise is so high that the linear segment finding is hampered. As a consequence, inaccurate corrections, which do not correspond to the last change of the interference amplitude, will cause errors (see the residual noise between the 11th and 14th s in Fig. 9).
Figure 9 Processed ECG signal+EMG noise+interference.
A very simple approach for suppression of EMG noise influence on the procedure introduces an additional parallel buffer where ongoing portions of the signal are unconditionally averaged [31]. This buffer is used for accurate linearity detection. Fig. 10 and Fig. 11 show the comparison of results without and with the parallel buffer. The traces 'a' represent ECG signal mixed with interference and EMG noise. The traces 'b' in Fig. 10 and 'c' in Fig. 11 indicate transitions from linear to non-linear segments (on-off functions). As can be seen, the averaged signal part in Fig. 10 is very limited. As a consequence, the procedure efficiency is reduced (Fig. 10c and 10d. In contrast, the parallel buffer allows detection of long linear segments (Fig. 11c and the residual noise in the processed signal (Fig. 11d is low. However, the noise is not totally suppressed because a part of it participates in the correction calculation.
Figure 10 Subtraction of interference+EMG noise without parallel buffer.
Figure 11 Subtraction of interference+EMG noise with parallel buffer.
Further improvement in EMG noise suppression is obtained by Christov [29] by using adaptive threshold value M, which is calculated with respect to the noise/signal ratio Rt, defined as Rt = SNL/SF, where the noise level SNL equals the summary width of the non-linear segments in an epoch SF, approximately corresponding to the RR interval under consideration. Linearity search with a criterion of M = 150 μV for 'clean' ECG (Fig. 12a) and for the same signal, but contaminated with EMG noise (Fig. 12b) show different SNL, (Fig. 12c and 12d). The method is developed in MATLAB environment. The initial threshold M is chosen equal to 30 μV. Then, it is gradually increased until Rt reaches 10%, after which the subtraction procedure is started. The level Rt ≤ 0.1 value is suggested by the ratio 'QRS width versus its RR interval', which is usually around 10%. The elimination of both interference and EMG noise can be observed in Fig. 13b and 13d, where M = 420 μV is used. For comparison, the 'clean' ECG signal is processed with M = 35 μV (Fig. 13a and 13c).
Figure 12 Noise levels SNL (traces c and d) in 'clean' ECG signal (trace a) and EMG contaminated signal (trace b), respectively.
Figure 13 Interference+EMG noise suppression using adaptive threshold M: 'clean signal' (a) processed with M = 35 μV (c) and contaminated signal (b) subjected to the procedure with M = 420 μV (d).
Linear segments cannot be regularly found in patients with atrial and ventricular fibrillation. However, the total preservation of the wave shape is not necessary for fibrillation detection and therefore, all kinds of traditional filters may be applied.
Interference suppression in high-resolution ECG
The subtraction procedure is not directly applicable to the body-surface His-ECG, as the low amplitude and relatively low frequency His-wave can not be distinguished in linear segments. Thus, the His-wave will be, in practice, suppressed or even removed from the signal. The EMG noise is usually of higher amplitude and with much higher frequency content compared to the surface His-wave. Therefore, simple change of the threshold value, M, does not result in acceptable delineation of linear and nonlinear segments.
Bazhina et al. [45,46] implemented the following modification. The beginning of the detected non-linear segment before a QRS complex is shifted 100 ms to the left, thus defining the His-wave region as a non-linear segment by default (Fig. 14).
Figure 14 The beginning of a non-linear segment including the QRS complex was shifted by default 100 ms to the left, to include the zone where the His-bundle wave is expected to appear.
The subtraction procedure and three other methods: notch filters, spectral interpolation [47], and regression subtraction [35] are tested against minimal distortion of the original signal [45]. The subtraction and the regression-subtraction procedures proved to be the best, as Baratta et al. [35] use a similar concept for noise estimation in linear segments. Regression-subtraction deals poorly with amplitude changes of the interference within the current segment.
Case of battery-supplied devices and computer aided ECG systems
The hardware measurement of fPL, necessary for compensation of the interference frequency modulation, is not feasible in battery-supplied devices and in some computer aided ECG systems. Dotsinsky and Stoyanov [12] studied the range of frequency changes of interference with constant amplitude, for which the residual part is restricted to acceptable levels without use of synchronized sampling. They found that residual interference below 20 μVp-p could be obtained with the procedure by: i) interference amplitude ≤ 0.4 mVp-p and ii) frequency change with a rate ≤ 0.0125 Hzs-1. Since such requirements for the power-supply can often be exceeded, a software interference measuring was developed.
The ECG signal is processed initially by a 49–51 Hz band-pass filter. The amplitudes of two adjacent samples, BrL and BrR, taken from a positive-going slope of the interference, located below and above the zero line, are measured (Fig. 15). The distance, tCP, between the crossing point CP and the right sample, BrR, is computed continually by:
Figure 15 Interference zero crossing.
In case of TPL change, tS is redefined using
This approach was implemented in the MATLAB environment. For sampling frequency fS = 500 Hz and fRPL = 50 Hz, n is equal to 10. The product kn determines the time recommended to elapse before calculating and substituting new tCP,i + kn for the previous tCP,i. Fig. 16 shows a processed 1 mV ECG signal after being mixed by interference with 2 mVp-p constant amplitude and extremely fast varying by 1 Hz per 8 s frequency (first trace). To assess the efficiency obtained, the zoomed difference without synchronized sampling (last trace) is also presented.
Figure 16 Subtraction procedure using software power-line frequency measurement.
The next logical step to be taken consists of: i) keeping the rated tS of the ECG instrument, ii) re-sampling the signal according to the ongoing measured fPL in order to eliminate the interference and iii) returning to the rated tS. The first results of such an approach are highly promising [48]. Thus, the software compensation of the variable fPL, as well as a total implementation of the subtraction procedure in an instrument, including automatic adjustment for fRPL of 50 or 60 Hz, will be completed regardless of the hardware circuits and the corresponding software.
Automatic adaptation to the rated PL frequency
A common program for alternative interference subtraction in 50 and 60 Hz environment leads to non-multiple sampling, i.e. to real n. Widely used values of tS for fRPL = 50 Hz, such as 250, 500 and 1000 Hz, correspond to irrational n of 4.1(6), 8.3(3) and 4.1(6) if 60 Hz interference has to be eliminated. In the inverse case, fS = 360 Hz requires n = 7.2. Rounded values n* are unacceptable to use, since they would introduce considerable error.
A very simple solution not needing fS change was found by Dotsinsky and Stoyanov [30]. The original procedure applies a comb filter over one period, TPL, of the interference. Thus, the program runs faster. Generally, n may be taken from k > 1 entire periods. The procedure is operated if:
n = kTPL/tS is an integer.
For tS = 2 ms (fS = 500 Hz) and fRPL = 50 Hz, the smallest value of k satisfying equation (14) is really k = 1. However, in case of 60 Hz, k equals 3. Zeros associated with the sub-harmonics 20 and 40 Hz will appear too but they have no influence on the procedure. Therefore, it is quite enough to switch n between 10 (k = 1) and 25 (k = 3) in order to operate with both interferences. For this purpose two digital band-pass filters check the incoming signal. Fig. 17 shows that the filter with frequency band overlapping the interference generates an order of magnitude higher output signal than the other filter.
Figure 17 Detection of the rated power-line frequency, a) part of QRS complex, b) 50 Hz interference and c) 60 Hz interference.
Theoretical procedure development
The theory of the subtraction procedure was developed further by Mihov [27], Levkov and Mihov [28], and Mihov et al. [32]. They proposed four types of filters, implemented in a generalized structure that may overcome the problems with almost all cases of non-multiple sampling, including interference frequency variations, without using synchronized AD conversion.
The so-called D-filter in multiple sampling is defined as is Cr in equation (2), where the second difference, Di, is obtained with FDs that are spaced at one TPL:
Di = (Xi+n - Xi) - (Xi - Xi-n) = Xi-n - 2Xi + Xi+n (15)
The transfer function of the D-filter has zeros at f = 0 and f = fPL Hz, which is shown equal to 50 Hz in Fig. 18.
Figure 18 Transfer function of D-filter for fPL = 50 Hz.
The K-filter describes the moving average presented by equations (5) and (6). Its transfer function is given in Fig. 19 for n = 5 in case of odd multiplicity.
Figure 19 Transfer function of K-filter for fPL = 50 Hz and n = 5.
The equation used for ongoing calculation of the interference components:
Bi = Xi - Yi (16)
actually defines a digital filter called (1-K)-filter.
Furthermore, the filters are re-defined for non-multiple sampling, and fS = 250 Hz in conjunction with fRPL = 60 Hz is taken in consideration to illustrate the software improvement.
In order to preserve the transfer function zeros, the D-filter has to be subtracted with a correction filter with zero at f = 0 and gain of DRPL at f = fRPL, equal to the gain of the D-filter for the same frequency, fRPL. The correction filter synthesis is based on a three-points auxiliary filter given by the equation:
where (n/2)* is the rounded value of n/2.
Since ARPL is the gain of the auxiliary filter for f = fRPL, the correction filter is multiplied by the ratio DRPL/ARPL. Using the corresponding transfer functions, DRPL and ARPL are computed in advance by:
Finally, the corrected D*-filter is presented as
and is shown in Fig. 20 by trace 'c', where traces 'a' and 'b' are the D-filter and the correction filter, respectively.
Figure 20 Transfer functions of a) D-filter, b) auxiliary correction filter and c) corrected D-filter.
The transfer function of the K-filter must preserve zero for f = fRPL, unity gain for f = 0 and linear phase response. The procedure of the K-filter correction is similar to the previous one. An auxiliary filter is given by the formula used for corrections computation:
Ai = Xi - Yi, (20)
The filter gain is equal to 1 - KRPL for f = fRPL, where KRPL is the K-filter gain for the same frequency fRPL. The auxiliary filter is multiplied by KRPL/(1 - KRPL) and subtracted from the K-filter. The equation for the corrected K*-filter is:
The constant KRPL can be estimated by:
for odd or even multiplicity, respectively.
An example of K-filter correction is shown in Fig. 21, where traces 'a', 'b' and 'c' represent the primary K-filter, the auxiliary filter, and the corrected K*-filter.
Figure 21 Transfer function of a) K-filter, b) auxiliary correction filter and c) corrected K-filter.
In case of non-multiple sampling, a phase difference appears between the ongoing ECG samples and the interference components Bi (equation 16) usually located in a temporary first-in-first-out (FIFO) buffer. Therefore, Bi must be modified being subtracted from the ECG samples during non-linear segments. The compensation procedure is relatively complicated. Fig. 22 shows the contents of the temporary buffer. The current interference sample, Bi-n*, does not coincide with the restored sample, Bi. Its amplitude must be recalculated in order to compensate the phase difference between them. This is accomplished by a new filter with linear phase response and unity gain for f = fRPL, denoted as the B-filter. It is synthesized from the known K-filter, with a window equal to the interference period. In case of odd n*, it can be described as:
Figure 22 Restored values in the temporary buffer.
where KRPL is the gain for the interference of the averaging filter given by equation (22).
The restored buffer value Bi can be calculated by:
In case of even n*:
The B-filter transfer function is shown in Fig. 23.
Figure 23 Transfer functions of a) B-filter and b) the known K-filter.
The generalized structure is presented in Fig. 24, where the modules of the subtraction procedure are as follows:
Figure 24 Generalized structure of the subtraction procedure.
• Linearity detection. D-filter is applied to evaluate the linearity of each signal sample neighbourhood.
• Interference extraction. (1-K)-filter is used to calculate the interference component.
• Criterion. The condition Cr <M sends either extracted or restored PL interference to Subtraction.
• Interference temporary buffer. The extracted or restored interference component used as correction in non-linear segment is saved at the position locked with the ongoing phase of the power-line interference.
• Interference restoring. B-filter is called in case of non-multiple sampling in order to restore the true correction values, which have to be subtracted from the input signal samples in non-linear segments.
• Delay buffer. Compensates the delay, which appears with the D-filter and (1-K)-filter and is imperative if the procedure is run in quasi-real time. Otherwise, the buffer could be disregarded.
• Subtraction. Extracted or restored interference value is subtracted from the delayed input signal to output 'clean' ECG signal. In case of non-linearity both Interference extraction and Subtraction implement the K-filter.
An improved algorithm according to the generalized structure has been tested off-line. The results for fS = 250 Hz and fRPL = 60 Hz are shown in Fig. 25.
Figure 25 Example for non-multiple sampling with fRPL = 60 Hz and fS = 250 Hz.
Conclusion
As first elaborated two decades ago and continually improved since then, the subtraction procedure eliminates power-line interference from the ECG signal without affecting its spectrum. The procedure operates successfully even with amplitude and frequency deviations of the interference. The frequency deviations are first compensated by hardware measurement of the power-line frequency. Software measurement of the interference period was developed for battery supplied units and some ECG modules connected to personal computers.
The improved structure of on-line going subtraction procedure leads to its extended implementation regardless of multiplicity between sampling rate and interference frequency. The structure flexibility is due to the introduced filtering modules, which are called into use depending on the type of sampling.
The presented analysis of the subtraction procedure and the different types of notch filters confirms the advantages of this method for interference cancellation in ECG signals.
Acknowledgements
The authors gratefully acknowledge the contribution of the Ph.D. students Mrs. Ts. Georgieva and Mr. T. Stoyanov in the theoretical investigations and the software synchronized sampling rate.
==== Refs
Huhta JC Webster JG 60 Hz interference in electrocardiography IEEE Trans Biomed Eng 1973 20 91 100 4688314
Thakor NV Webster JG Ground free ECG recording with two electrodes IEEE Trans Biomed Eng 1980 27 699 704 7461645
Towe BC Comments on 'Ground-Free ECG Recording with Two Electrodes' IEEE Trans Biomed Eng 1981 28 838 839 7343473
Metting van Rijn Peper A Grimbergen CA High-quality recording of bioelectrical events, Part 1: Interference reduction, theory and practice Med Biol Eng Comput 1990 28 389 397 2277538
Pallas-Areny R Interference-rejection potential characteristics of biopotential amplifiers: A comparative analysis IEEE Trans Biomed Eng 1988 35 953 959 3198140
Metting van Rijn Peper A Grimbergen CA The isolation mode rejection ratio in Bioelectric amplifiers IEEE Trans Biomed Eng 1991 38 1154 1157 1748450
Sahambi JS Tandon SN Bhatt RKP Quantitative analysis of errors due to power line interference and base line drift in detection of onsets and offset in ECG using wavelets Med Biol Eng Comput 1997 35 747 751 9538556
Pei SC Tseng CC Elimination of AC Interference in Electrocardiogram Using IIR Notch Filter with Transient Suppression IEEE Trans Biomed Eng 1995 42 1128 1132 7498917
Ma WK Zhang YT Yang FS A fast recursive-least-squares adaptive notch filter and it is applications to biomedical signals Med Biol Eng Comput 1999 37 99 103 10396849
Hamilton PS A comparison of Adaptive and Nonadaptive Filters for Reduction of Power Line Interference in the ECG IEEE Trans Biomed Eng 1996 43 105 109 8567001
Yoo SK Kim NH Song JS Lee TH Kim KM Simple self tuned notch filter in a bio-potential amplifier Med Biol Eng Comput 1997 35 151 154 9136210
Dotsinsky I Stoyanov T Power-line interference cancellation in ECG signals Biomed Instr Techn 2005 39 155 162 15810791
Dotsinsky IA Daskalov IK Accuracy of the 50 Hz Interference Subtraction from the Electrocardiogram Med Biol Eng Comput 1996 34 489 494 9039755
Mitov IP A method for reduction of power line interference in the ECG Med Eng Phys 2004 26 879 887 15567704
Bellanger M Digital Processing of Signals: Theory and Practice 2000
Thakor NV Zhu Y Applications of adaptive filtering to ECG analysis: noise cancellation and arrhythmia detection IEEE Trans Biomed Eng 1991 38 785 793 1937512
Bensadoun Y Raoof K Novakov E Elimination du 50 Hz du signal ECG par filtrage adaptatif multidimensionnel Innov Tech Biol 1994 15 751 758
Kumaravel N Nithiyanandam N Genetic-algorithm cancellation of sinusoidal powerline interference in electrocardiograms Med Biol Eng Comput 1998 36 191 196 9684459
De Lima LA Yioneyama T A neural integrated circuit for noise cancellation in electrocardiograph signals Proceedings of the 11th International Conference on Control Systems and Computer Science: Bucharest 1997 101 105 26–28 May 1997
Romanca M Szabo W Electrocardiogram pre-processing for the removal of high frequency and power-line frequency noise Proceedings of the 6th International Conference on Optimization of Electrical and Electronic Equipments: May 1998; Braşov 1998 703 706
Fergjallah M Barr RE Frequency-domain digital filtering techniques for the removal of power-line noise with application to the electrocardiogram Comput Biomed Res 1990 23 473 489 2225791
Levkov C Michov G Ivanov R Daskalov I Subtraction of 50 Hz interference from the electrocardiogram Med Biol Eng Comput 1984 22 371 373 6748774
Christov II Dotsinsky IA New approach to the digital elimination of 50 Hz interference from the electrocardiogram Med Biol Eng Comput 1988 26 431 434 3255855
Yan XG Dynamic Levkov-Christov subtraction of mains interference Med Biol Eng Comput 1993 31 635 638 8145592
Dotsinsky IA Daskalov IK Comments on 'Dynamic Levkov-Christov subtraction of mains interference' Med Biol Eng Comput 1995 33 360 7475379
Daskalov I Dotsinsky I Reader respond to 'A new adaptive coherent model algorithm for removal of power-line interference' J Clinical Eng 1995 20 357 358
Mihov G Jan J, Kilian P, Provaznik I Elimination of mains interference from the ECG in non-synchronized sampling: a theoretical approach Proceedings of the 13th Biennial International Conference BIOSIGNAL'96: June 1996; Brno 1996 Technical University Brno Press 189 191
Levkov C Mihov G Jan J, Kilian P, Provaznik I Rejection-subtraction filter of mains interference from the ECG Proceedings of the 13th Biennial International Conference BIOSIGNAL'96: June 1996; Brno 1996 Technical University Brno Press 183 185
Christov II Dynamic powerline interference subtraction from biosignals J Med Eng Techn 2000 24 169 172 11105290
Dotsinsky I Stoyanov T Cancellation of the power-line interference: effect of amplitude and frequency variations on the ECG signal Proceedings of the 11-th International Conference ELECTRONICS – ET'20: Sozopol 2002 65 70 25–27 September 2000
Dotsinsky I Christov I Power-line interference subtraction from the electrocardiogram in the presence of electromyogram artifacts Electrotechniques & Electronics E+E 2002 1–2 18 21
Mihov G Dotsinsky I Georgieva Ts Subtraction procedure for power-line interference removing from ECG: Improvement for non-multiple sampling J Med Eng Techn 2005 29 238 243
Dotsinsky IA Christov II Levkov CL Daskalov IK A microprocessor-electrocardiograph Med Biol Eng Comput 1985 23 209 212 4021565
Daskalov IK Dotsinsky IA Christov II Developments in ECG Acquisition, Preprocessing, Parameter Measurement and Recording IEEE Eng Med Biol 1998 17 50 58 9548081
Baratta RV Solomonov M Zhou BH Zhu M Methods to reduce the variability of EMG power spectrum estimates J Electromyography Kinesiology 1998 8 279 285 9785248
Li G Ling L Qilian Y Xuemin Y A new adaptive coherent model algorithm for removal of power-line interference J Clin Eng 1995 20 147 150
Wu Y Yang Y A new digital filter method for eliminating 50 Hz interference from ECG Chinese J Med Instrum 1999 23 145 158
Monaco A Sviluppo di un modulo software per la gestione di un sistema per il controllo remoto dei portatori di pacemaker 2000
Adli Yamamoto Y Nakamura T Kitaoka K Automatic interference controller device for eliminating the power-line interference in biopotential signals Proceedings of the 17th IEEE Instrumentation and Measurements Technology Conference: Baltimore 2000 1358 1362 01–04 May 2000
Ziarani AK Konrad A A Nonlinear Adaptive Method of Elimination of Power Line Interference in ECG Signals IEEE Trans Biomed Eng 2002 49 540 547 12046699
Butler KE Russell RD Subtraction of power-line harmonics from geophysical records Geophysics 2003 58 898 903
Lynn PA Online digital filters for biological signals: some fast designs for a small computer Med Biol Eng Comput 1977 15 534 540 199808
Kumaravel N Senthil A Sridhar KS Nithiyanandam N Integrating the ECG power-line interference removal methods with rule-based system Biomed Sci Instrum 1995 31 115 120 7654947
McManus CD Neuber KD Cramer E Characterization and elimination of AC noise i electrocardiograms: a comparison of digital filterimg methods Comput Biomed Res 1993 26 48 67 8444027
Bazhyna A Christov I Gotchev A Daskalov I Egiazarian K Powerline Interference Suppression in High-Resolution ECG Comp in Card 2003 30 561 564
Bazhyna A Christov I Gotchev A Daskalov I Egiazarian K Beat-to-beat noise removal in noninvasive His-bundle electrocardiogram Med Biol Eng Comput 2004 42 712 720 15503974
Mewett DT Nazeran H Reynolds KJ Removing power line noise from recorded EMG Proceedings of the 23rd Annual International Conference on Engineering in Medical and Biological Society: Istanbul 2001 2190 2193 25–28 October 2001
Dotsinsky I Removal of frequency fluctuating power-line interference from ECG Proceedings of the 3rd European Medical & Biological Engineering Conference: Prague 20–25 November 2005. Accepted
|
16117827
|
PMC1232857
|
CC BY
|
2021-01-04 16:37:34
|
no
|
Biomed Eng Online. 2005 Aug 23; 4:50
|
utf-8
|
Biomed Eng Online
| 2,005 |
10.1186/1475-925X-4-50
|
oa_comm
|
==== Front
J CarcinogJournal of Carcinogenesis1477-3163BioMed Central London 1477-3163-4-141613525710.1186/1477-3163-4-14ResearchMucin phenotypic expression and p53 gene abnormality of gastric super-minute well-differentiated adenocarcinoma: Re-evaluation with relationship between histogenesis of well-differentiated adenocarcinoma and intestinal metaplasia in distal stomach Wada Ryo [email protected] Toshikazu [email protected] Takayuki [email protected] The Department of Pathology, Juntendo Shizuoka Hospital of Juntendo University School of Medicine, Shizuoka, Japan2 The Department of Pathology(I), Juntendo University School of Medicine, Tokyo, Japan3 R & D Center, Biomedical Laboratories, Inc, Saitama, Japan2005 1 9 2005 4 14 14 16 3 2005 1 9 2005 Copyright © 2005 Wada et al; licensee BioMed Central Ltd.2005Wada 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
Although the gastric well-differentiated adenocarcinoma in the distal stomach has been thought to develop via a intestinal metaplasia-carcinoma sequence, there are some disproofs from new mucin examinations for minute-size lesions in same type carcinoma. The current study was performed and pointed out the new findings for the solution to the problem according to the point described above.
Methods
12 super-minute lesions (less than 1 mm in maximum diameter) of well-differentiated adenocarcinoma in distal stomach (SMCa), which were detected from the pathological examinations of 210 surgically resected stomach specimens, and the mucosa adjacent to these carcinoma lesions, were examined by immunohistochemical mucin stainings (MUC2 and CD-10: intestinal phenotype, 45M1 and MUC6: gastric phenotype) and p53-overexpression. And the analyses of the replication error of the microsatellites in chromosome 17 related p53 gene (TP53 and D17S786) (RER-p53MS) were performed in SMCa lesions, adjacent mucosa to each lesion and other gastric mucosa with intestinal metaplasia, because all SMCa lesions showed p53-overexpression immunohistochemically, decribed below.
Results
1. The carcinoma cells in all SMCa lesions were positive for 45M1 and p53. On the other hand, no positive carcinoma cells for MUC6 were seen although the pyloric glands and the remnant pyloric gland in the SMCa lesions in the same slides were positive for MUC6. Ten lesions (83%) had intestinal phenotypic mucin (10 lesions: MUC2 (+), 4 lesions: CD10 (+)). Two lesions (17%) were positive for only 45M1 (gastric phenotypic mucin). 2. All of the mucosa adjacent to SMCa showed intestinal metaplasia (complete type: 7 regions, incomplete type: 5 regions). 3. RER-p53MS was confirmed in 42% (5/12 regions) of SMCa, in 42% (5/12 regions) of the mucosa adjacent to SMCa and 14% (6/42 regions) of the other intestinal metaplasia mucosa.
Conclusion
Most of the super-minute well-differentiated adenocarcinoma lesions in the distal stomach, which had both gastric and intestinal phenotypic mucin, are considered to develop from the tubular proliferative zone with the incomplete type of the intestinal metaplasia and p53 gene abnormality, while a part of them, which had only gastric phenotypic mucin, may derive from the gastric native tubules (non-metaplastic epithelium) with p53 gene abnormality.
Gastric differentiated adenocarcinomamucinmicrosatellitep53MUC2CD1045M1MUC6
==== Body
Background
Although the gastric differentiated adenocarcinoma has been thought to develop via a intestinal metaplasia-carcinoma sequence [1-4], there are some disproofs [5] that the gastric differentiated adenocarcinoma develops having a poor relationship to the intestinal metaplasia, from new mucin examinations for minute-size carcinoma [6,7] in same type. Thus, some problems for that histogenesis of the gastric differentiated adenocarcinomaare considered to be still remained.
To clarify the true histogenesis of digestive tract carcinoma, the extremely small-sized carcinoma lesions were found and these lesions including the mucosa adjacent to them must be examined pathologically and molecular-biologically. Because almost all lesions of the digestive tract carcinoma develop from the mucosal epithelium and the minute-sized carcinoma lesions and the mucosa adjacent to them may express the initial condition when each carcinoma develop.
According to the point described above, the current study was performed and pointed out the new findings for the solution to this problem.
In this study, firstly, the immunohistochemical mucin-stainings and p53-overexpression, and secondary, the replication error (RER) of the microsatellite (MS) in chromosome 17 related p53 gene of the super-minute sized differentiated adenocarcinoma lesions in the human stomachs, whose definition was described below, were investigated, because the carcinoma cells were positive p53 in all carcinoma lesions in the current study, and the new findings on histogenesis of gastric adenocarcinoma were revealed.
Methods
Twelve tubular proliferative lesions (within 1 mm in maximum diameter) with cellular atypia and the boundaries to the neighborhood, which were diagnosed as high-grade intraepithelial neoplasia according to both the Vienna and WHO classification [8,9] and as well-differentiated adenocarcinoma according to the classification of the Japanese Research Society for Gastric Cancer [10] (the super minute-sized well-differentiated adenocarcinoma in the distal stomach: SMCa) (Figure 1 &2), were detected in the distal stomachs from the pathological examinations of 210 surgically resected stomach specimens with early cancer, assessed at the Department of Pathology, Juntendo University hospital and Shizuoka Hospital, between 1982 and 2003. Although in the cases between 1999 and 2003 the informed consent for pathological and molecular examinations were obtainted from each patient, in other cases the informed consent could not be obtained because the system of the informed consent for patients was built up at 1999.
Figure 1 Typical case of SMCa. SMCa lesion is seen in the center region of this fugure. (×20, HE)
Figure 2 High power view of Figure 1 lesion. This lesion was diagnosed as high-grade intraepithelial neoplasia according to both the Vienna and WHO classification and as well-differentiated adenocarcinoma according to the classification of the Japanese Research Society for Gastric Cancer. (×200, HE)
The stomach specimens described above were fixed in 10% buffered formalin solution and prepared by cutting from almost the entirety of the stomachs area into 3 – 5 mm wide sections. Each section was embedded in paraffin and stained with hematoxylin and eosin (HE). And every sections with 12 SMCa lesions described above were cut serially and stained with HE, and the greatest dimensions of these lesions were determined (655 – 958 μm). The cases with SMCa lesions were considered as the cases of multiple gastric cancers by the criteria of Moertel et al. [11].
Using all paraffin blocks with SMCa lesions and 10 paraffin blocks with no neoplasia in each stomach, and the mucosa adjacent to SMCa lesions were examined by immunohistochemical mucin stainings as shown in Table 1 (MUC2 and CD-10: intestinal phenotype, 45M1 and MUC6: gastric phenotype). These 4 mucin stainings are more popular for detection the mucin phenotypes of the gastric differentiated adenocarcinoma [12].
Table 1 Immunohistochemical mucin antibodies
Commercial name Positive cell Dilution Source
MUC2 Human Muc2 glycoprotein Goblet cell and their precursor cell 1:500 Novocastra, UK
CD10 CD10 glycoprotein Enterocytes 1:200 Novocastra, UK
45M1 Peptide core of human gastric mucin (NCL-HGM-45M1) Gastric surface mucous cell 1:100 Novocastra, UK
MUC6 Human Muc6 glycoprotein Pyloric gland, mucous neck cell, cardiac gland, Brunner's gland 1:100 Novocastra, UK
And we have tried to examine a number of Exons for detecting the p53 gene abnormalities, because all lesions of the gastric super-minute well-differentiated adenocarcinomas showed overexpression of p53 immunohistochemically, described below (in "Results"). However, it was hard to examine it, because the target fici were too small and in post-formalin solution condition. Secondary, we have tried to estimate the existence or nonexistence of the p53 gene abnormalities using microsatellite markers in chromosome 17 related p53 gene (RER-p53MS).
Immunohistochemical stainings
Immunohistochemical mucin-stainings were performed by the avidin-biotin-peroxidase-complex method with each condition as shown in Table 1. Recently, it has been pointed out that the data of the mucin phenotypes of the lesions will change accroding to what percentage of positive cells in each stainings is judged as the positive in each stainings [13]. In this study, the mucin phenotypes of the lesions were determined by positive (>0% of the carcinoma cells) or negative. Because the target lesions in this study were extremely small lesions, and even if only a carcinoma cell was positive, this positivity should be considered to be the effective and valuable data. And p53-oncoprotein antibody staining with microwave treatment was done in many sections (p53: DO-7, Novocastra Inc., UK).
DNA extraction and analysis of RER-p53MS using microsatellite markers (TP53 and D17S786)
Paraffin blocks with the target foci mentioned above were prepared for DNA extraction. The target foci were microdissected using a 20-gauge needle, comparing the slide with HE staining in the same position. The extracted DNA was diluted with 500 μl DNA Isolation Solution (DNA isolater for DNA Extraction from Paraffin-embedded Tissue, Wako Inc., Osaka). In this study, target foci were SMCa and other gastric non-neoplastic mucosa and the extracted DNA from them were used as template DNA in polymerase chain reaction (PCR). However, in approximately half of target foci, DNA could not be extracted, because these target foci may be too small and moreover these materials were done by the formalin-solution. And decribed above, as the purpose of the molecular biological examinations was to know the existence or nonexistence of p53 gene abnormalities of target foci, TP53 and D17S786 were selected as microsatellite markers in the current study, according the proposition by National Cancer Institute Workshop [14].
RER-p53MS was investigated using high resolution fluorescenced labeled PCR primers in proportion to the method of Tsuchida et al. [15]. The outline of it: 1. PCR was performed containing 1 μL of DNA lysate, 100 μM dNTP, 1.5 mM MgCl2, 1 μM each primer marked with fluorescent dye of three colors as blue, green and yellow, 0.625 U Taq DNA polymerase and 1 × PCR buffer [containing 10 mM Tris-HCl (pH 8.3 at 25°C), 50 mM KCl and 0.001%(w/v) gelatin] in a thermal cycler. 2. The electrophoresis was conducted 2 hours by means of an ABI-377 DNA auto-sequencer (PE Bio-systems, Inc., Foster City, CA, USA). 3. A comparison was made of peaks of same marker arising from gastric non-neoplastic and metaplastic mucosa and SMCa lesions, using the Gene Scan TM waveform analyzed softwave, and shift were assessed. The shift (+) was determined when a new peak not recognized with non-neoplastic and metaplastic tissue was confirmed with the target mucosa with intestinal metaplasia or SMCa lesions, that is to say, RER-p53MS was positive (Figure 3).
Figure 3 The shift (+) was determined when a new peak not recognized with non-neoplastic and metaplastic tissue (lower view in Figure 3) was confirmed with the target mucosa with intestinal metaplasia or SMCa lesions (upper view in Figure 3), that is to say, RER-p53MS was positive.
Although we have tried to examine the existence of Helicobacter pylori in the SMCa lesions and the mucosa adjacent to them, the surface mucous layer of specimen stomachs, in which Helicobacter pylori may often exist, was washed away at taking photograph of each specimen, and the relationship between SMCa and Helicobacter pylori could not be examined.
The data were analyzed statistically with Student's t-test (t-test) and chi-square test; a p-value of less than 0.05 was considered to be significant.
Results
1. Mucin phenotypes of SMCa (Table 2)
Table 2 Mucin phenotypes of SMCa
Case MUC6 45M1 MUC2 CD10
1 - + + +
2 - + - -
3 - + + -
4 - + + +
5 - + + -
6 - + + -
7 - + + -
8 - + - -
9 - + + +
10 - + + -
11 - + + -
12 - + + +
Total 0% (0/12) 100% (12/12) 83% (10/12) 34% (4/12)
The carcinoma cells in all SMCa lesions were positive for 45M1 (Figure 4). On the other hand, no carcinoma cells in all SMCa lesions were positive for MUC6, although the pyloric glands and the remnant pyloric glands in SMCa lesions, which were seen in the same slides, were positive for MUC6. Ten lesions (83%) had intestinal phenotypic mucin (10 lesions: MUC2 (+), 4 lesions: CD10 (+)) (Figure 5). Two lesions (17%) were positive for only 45M1. That is to say, among SMCa lesions, 10 lesions (83%) had both gastric and intestinal phenotypic mucin and 2 lesions (17%) had only gastric phenotypic mucin.
Figure 4 Among SMCa lesions, all lesions were positive for 45M1. (×200, 45M1 in Figure 1 lesion)
Figure 5 Ten lesions were positive for MUC2. (×200, MUC2 in Figure 1 lesion)
2. The mucosa adjacent to SMCa lesions
All of the mucosa adjacent to SMCa showed intestinal metaplasia (complete type: 7 regions, incomplete type: 5 regions, by the criteria of Filipe et al. [16]).
3. p53-overexpression and frequency of RER-p53MS in SMCa and the mucosa adjacent to them
Although most of the carcinoma cells in every SMCa was positive for p53 (Figure 6), there was no positive cell for p53 in the mucosa adjacent to SMCa and other intestinal metaplasia.
Figure 6 Anti-p53 antibody staining (p53) in Figure 1 lesion. Most of the carcinoma cells in this lesion showed a positive reaction for p53. (×200, p53)
RER-p53MS was confirmed in 42% (5/12 regions) of SMCa, in 42% (5/12) of the mucosa adjacent to SMCa and 14% (6/42) of the other intestinal metaplasia mucosa (complete type: 17%, 4/23, incomplete type: 11%, 2/19), as shown in Table 3.
Table 3 Frequency of RER-p53MS in SMCa, the mucosa adjacent to them and intestinal metaplasia
TP53 D17S786
SMCa 42% (5/12 resions) 8% (1/12 regions)
Mucosa adjacent to SMCa (intestinal metaplasia) 42% (5/12 regions) 17% (2/12 regions)
Other intestinal metaplasia 14% (6/42 regions) 5% (2/42 regions)
Discussion
Although gastric differentiated adenocarcinoma is thought to develop from intestinal metaplasia [1-4], there are the disproof reports using a new mucin examination methods [5-7]. Thus, the histogenesis of the gastric differentiated adenocarcinoma may be still unclear.
The small carcinoma lesion is thought to be a good model for researching the histogenesis of digestive tract carcinoma, because almost all lesions of the digestive tract carcinoma develop from the mucosal epithelium and the minute-sized carcinoma lesions and the mucosa adjacent to them may express the initial condition when each carcinoma develop. Therefore, the carcinoma lesions must be as small as possible to examine the original findings in the mucin phenotypes or the gene abnormalities, and moreover the essential mucin expression in the small lesions must be determined as 'positive' even if a carcinoma cell was positive.
Thus, the mucin expression and gene abnormalities in super-minute-sized gastric well-differentiated adenocarcinoma, which have been shown in the initial stages of same type-carcinoma, are still thought to be unclear and the current study was performed focussing on this, and some interesting findings were revealed in the current study.
SMCa in the current study was defined as the tubular proliferation with cellular and structural atypia, diagnosed as high-grade intraepithelial neoplasia according to the Vienna [8] and WHO classification [9], and as well-differentiated adenocarcinoma according to the classification of the Japanese Research Society for Gastric Cancer [10]. Thus, SMCa lesions were thought to be the initial stage of development of the gastric well-differentiated adenocarcinoma.
Our results in the current study showed that 1. Among SMCa lesions, all carcinoma lesions were positive for 45M1 and p53, and no carcinoma cells were positive for MUC6. Ten SMCa lesions (83%) had intestinal phenotypic mucin (10 lesions: MUC2 (+), 4 lesions: CD10 (+)). Two SMCa lesions (17%) were positive for only 45M1. That is to say, among SMCa lesions, 10 lesions (83%) had both gastric and intestinal phenotypic mucin and 2 lesions (17%) had only gastric phenotypic mucin. 2. All of the mucosa adjacent to SMCa showed intestinal metaplasia (complete type: 7 regions, incomplete type: 5 regions). 3. RER-p53MS was confirmed in 42% (5/12 regions) of SMCa, in 42% (5/12 regions) of the mucosa adjacent to SMCa and 14% (6/42 regions) of the other intestinal metaplasia mucosa.
Although the mucin phenotypes of SMCa lesions differed from the data of other reports using the gastric minute differentiated adenocarcinoma [6,7], the reason of the differences may be caused the differences of the definition of the positivity in each stainings decribed above. In the current study, the lesions were judged as 'positive' even if a carcinoma cell was positive in each staining. The data in RER-p53MS of SMCa was almostly equal to the data of MS instability or the loss of heterozygosity in the gastric differentiated adenocarcinoma and intestinal metaplasia in other reports [17-19].
Focusing on the histogenesis of the gastric diffrentiated adenocarcinoma, the results of mucin expressions of SMCa revealed that gastric well-differentiated adenocarcinoma should not derive from the intestinal metaplasia with no remnant gastric original tubules (the tubules exchanged perfectly by intestinal metaplasia), because all of SMCa lesions had the gastric phenotypic mucin. However, most of SMCa may develop having a relashionship to the intestinal metaplasia, because they had also the intestinal phenotypic mucin. Namely, most of the SMCa lesions should derive from the gastric native tubule with a part of intestinal metaplasia and p53 gene abnormality.
On the other hand, a minority of SMCa had only gastric phenotypic mucin, are considered to develop from the gastric native tubules with no intestinal metaplasia and with p53 gene abnormality.
Thus, when the judgements of the mucin expressions in the gastric super-minute well-differentiated adenocarcinoma were strictly performed like as the current study, although the main route of the histogenesis of the gastric differentiated adenocarcinoma is thought to be the intestinal metaplasia-carcinoma sequence as before, it is very important that some well-differentiated adenocarcinoma may derive from the gastric native tubules with no relationship to the intestinal metaplasia. That is say, the intestinal metaplasia should not be the indispensable factor in the development of the well-ifferentiated adenocarcinoma in the distal stomach.
In the current study, the target lesions were limited to the lesions in the distal stomach, because the carcinoma lesions as like SMCa lesions could not be found in proximal stomach including the gastric cardia. Thus, we could not point out the histogenesis of the gastric well-differentiated adenocarcinoma in the whole stomach. The further study between the histogenesis of gastric differentiated adenocarcinoma and the intestinal metaplasia is expected.
Abbreviations
SMCa, super-minute lesions (less than 1 mm in maximum diameter) of well-differentiated adenocarcinoma in distal stomach; RER, replication error; MS, microsatellite
Acknowledgements
The authors thank Mr. Mrozek D assistance with manuscript, and many people, begining Dr. Noriyuki Kuwabara (Juntendo University) and Dr. Koichi Suda (Juntendo University) for support in collecting the targets lesions, and Dr. Michio Matsumoto (Juntendo University Shizuoka Hospital) for support in doing this study.
==== Refs
Jarvi O Lauren P On the role of heterotopias of the intestinal epithelium in the pathogenesis of gastric cancer Acta Pathol Microbiol Scand 1951 29 26 43 14902458
Morson BC Carcinoma arising from areas of intestinal metaplasia in the gastric mucosa Br J Cancer 1955 9 377 385 13269635
Nakamura K Sugano H Takagi K Carcinoma of the stomach in incipient phase: its histogenesis and histological appearances Gann 1968 59 251 258 5726267
Correa P A human model of gastric carcinogenesis Cancer Res 1988 48 3554 3560 3288329
Tatematsu M Ichinose M Miki K Gastric and intestinal phenotypic expression of human stomach cancers as revealed by pepsinogen immunohistochemistry and mucin histochemistry Acta Pathol Jpn 1990 40 494 504 2220396
Egashira Y Shimoda T Ikegami M Mucin histochemical analysis of minute gastric differentiated adenocarcinoma Pathol Intern 1999 49 55 61 10.1046/j.1440-1827.1999.00824.x
Kawachi H Takizawa T Eishi Y Absence of either gastric or intestinal phenotype in microscopic differentiated gastric carcinomas J Pathol 2003 199 436 446 12635134 10.1002/path.1323
Schlemper RJ Riddle RH Kato Y The Vienna classification of gastrointestinal epithelial neoplasia Gut 2000 47 251 255 10896917 10.1136/gut.47.2.251
Stanley RH Aaltonen LA eds World Health Organization classification of tumours Pathology and genetics of tumours of the digestive system 2000 Lyon: IARC Press
Japanese Research Society for Gastric Cancer Japanese Classification of Gastric Cancer 1995 Tokyo: Kanehara
Moertel CG Bargen JA Soule EH Multiple gastric cancers: review of the literature and study of 42 cases Gastroenterol 1957 32 1095 1103
Tsukashita S Kushima R Bamba M MUC gene expression and histogenesis of adenocarcinoma of the stomach Int J Cancer 2001 94 166 170 11668493 10.1002/ijc.1460
Shiroshita H Watanabe H Ajioka Y Re-evalution of mucin phenotypes of gastric minute well differentiated-type adenocarcinoma using a series of HGM, MUC5AC, MUC6, M-GGMC, MUC2 and CD10 stains Pathol Intern 2004 54 311 321 10.1111/j.1440-1827.2004.01625.x
Boland CR Thibodeau SN Hamilton SR A National Cancer Insitute Workshop on microsatellite instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer Cancer Res 1998 58 5248 5257 9823339
Tsuchida A Aoki T Kasuya K Micro-satellite instability in multiple and single gastric cancers using fluorescent and auto-sequencer Ann Cancer Res Ther 2000 8 112 124
Filipe MI Munoz N Matko I Intestinal metaplasia types and the risk of gastric cancer: a cohort study in Slovenia Int J Cancer 1994 57 324 329 8168991
Hamamoto T Yokozaki H Semba S Altered microsatellites in incomplete-type intestinal metaplasia adjacent to primary gastric canccers J Clin Pathol 1997 50 841 846 9462267
Leung WK Kim JJ Kim JG Microsatellite instability in gastric intestinal metaplasia in patients with and without gastric cancer Am J Pathol 2000 156 537 543 10666383
Kobayashi K Okamoto T Takayama S Genetic instability in intestinal metaplasia is frequent event leading to well-differentiated early adenocarcinoma of the stomach Eur J Cancer 2000 36 1113 1119 10854944 10.1016/S0959-8049(00)00066-6
|
16135257
|
PMC1232858
|
CC BY
|
2021-01-04 16:39:20
|
no
|
J Carcinog. 2005 Sep 1; 4:14
|
utf-8
|
J Carcinog
| 2,005 |
10.1186/1477-3163-4-14
|
oa_comm
|
==== Front
J CarcinogJournal of Carcinogenesis1477-3163BioMed Central London 1477-3163-4-151613891810.1186/1477-3163-4-15ResearchApoptosis induced by the Tibetan herbal remedy PADMA 28 in the T cell-derived lymphocytic leukaemia cell line CEM-C7H2 Jenny Marcel [email protected] Wolfgang [email protected] David [email protected] Oliver A [email protected] Daria [email protected] Dietmar [email protected] Florian [email protected] Division of Medical Biochemistry, Biocenter, Innsbruck Medical School, Fritz Pregl-str. 3, 6020, Innsbruck, Austria2 Division of Experimental Pathophysiology and Immunology, Biocenter, Innsbruck Medical School, Fritz Pregl-str.3, 6020, Innsbruck, Austria3 Department of Oncology-Pathology, Karolinska Institute, Cancer Centrum R 8:00, 17176, Stockholm, Sweden4 Division of Biological Chemistry, Biocenter, Innsbruck Medical School, Fritz Pregl-Str. 3, 6020, Innsbruck, Austria.2005 2 9 2005 4 15 15 12 8 2004 2 9 2005 Copyright © 2005 Jenny et al; licensee BioMed Central Ltd.2005Jenny 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.
The Tibetan herbal remedy PADMA 28 revealed promising results to support treatment of atherosclerosis, Charot syndrome (intermittent claudication), chronic active hepatitis and infection of the respiratory tract. The remedy was confirmed to be closely linked with anti- and pro-oxidative properties in vitro. In this study, apoptogenic and survival effects of PADMA 28 were investigated in the T cell-derived lymphocytic leukaemia cell line CEM-C7H2.
PADMA 28 led to a concentration-dependent inhibition of cell proliferation accompanied by the accumulation of CEM-C7H2 cells in subG1 phase, fragmentation of poly (ADP-ribose) polymerase (PARP) and nuclear body formation. Treatment with PADMA 28 rescued to some extent cells over-expressing Bcl-2 from apoptosis. This finding suggests that the mechanism of action of PADMA 28 may be via interference with Bcl-2 triggered survival pathways.
ApoptosisBcl-2T cell-derived lymphocytic leukaemia cellsMulticomponent therapeutics, Tibetan medicinePADMA 28
==== Body
Background
In its composition, PADMA 28 largely corresponds to an herbal formula derived from Tibetan medicine that was used in Poland in the first half of the 20th century against chronic inflammatory diseases. In the past two decades, this formula has been administered with or without boiled water cooled to lukewarm hot water to a large number of patients suffering from peripheral vascular occlusive disease in Europe and revealed promising results in intermittent claudication [1], atherosclerosis [2], and chronic active hepatitis B [3].
It was reported to have both anti- and pro-oxidative properties in vitro [4]. In mouse macrophages PADMA 28 was found to inhibit inducible nitric oxide synthesis (iNOS) by decreasing iNOS mRNA and iNOS protein levels [5]. Recently, Neurauter et al. demonstrated that PADMA 28 modulates interferon γ-induced tryptophan degradation and neopterin production in human peripheral blood mononuclear cells (PBMC) [6]. Others have reported on its anti-inflammatory and anti-atherogenic properties [7].
It is generally accepted that improper functioning of cellular suicide or apoptosis is one of the master switches in tumourigenesis [8-10]. Accumulation of cells in subG1 phase, degradation of poly (ADP-ribose) polymerase (PARP) and nuclear body formation are hallmarks of apoptosis. In this study, using CEM-C7H2 leukaemia cells, we investigated whether and in what way PADMA 28 influences apoptosis.
Bcl-2 is the central gene family member involved in balancing between cell death and cell survival [11]. In the present study, the influence of PADMA 28 on Bcl-2 signalling in CEM-C7H2 cells deficient in functional p53 and p16 signalling was studied in a CEM sub-line constructed with the Bujard system in which the Bcl-2 gene is induced upon withdrawal of tetracycline [12].
Methods
Test substance
PADMA 28 was provided in powder form as an active substance mixture without excipients by PADMA Inc. Schwerzenbach, Switzerland. The herbal mixture is a fine brown dusty powder with a characteristic pungent smell. It is a mixture of a variety of different herbs (403 mg of active ingredients): Aegle marmelos (Bengal Quince) fruit (20 mg); Pimenta dioica (Allspice) fruit (25 mg); Aquilegia vulgaris (Columbine) aerial part (15 mg); Calendula officinalis (Marigold) flower (5 mg); Elettaria cardamomum (Cardamom) fruit (30 mg); Syzygium aromaticum (Clove) flower bud (12 mg); Saussurea lappa (Saussuria) root (40 mg); Hedychium spicatum (Hedychium) rhizome (10 mg); Lactuca sativa (Lettuce) leaf (6 mg); Cetraria islandica (Iceland moss) thallus (40 mg); Glycyrrhiza glabra (Licorice) root (15 mg); Azadirachta indica (Margosa) fruit (35 mg); Terminalia chebula (Myrobalan) fruit (30 mg); Plantago lanceolata (Ribwort) aerial part (15 mg); Polygonum aviculare (Knot-grass) aerial part (15 mg); Potentilla aurea (Golden Cinquefoil) aerial part (15 mg); Pterocarpus santalinus (Red Sandalwood) wood (30 mg); Sida cordifolia (Heart-leaved Sida) aerial part (10 mg); Valeriana officinalis (Valerian) root (10 mg); and Aconitum napellus (Monkshood) tuber (1 mg). Also present are 2 non-herbal components: Dextrocamphora (natural camphor) (4 mg) and Calcii sulphas pulv.(Gipsum; 20 mg).
In parallel, a derivative of this mixture, PADMA Basic, was also studied and its effects compared to PADMA 28. In view of different pharmaceutical standards in Austria and United States, Aconitum napellus is omitted in PADMA Basic.
Since the data on PADMA 28 were fully concordant with those on PADMA Basic only figures showing PADMA 28 data are shown.
Individual constituents as well as the whole mixtures were inspected and manufactured according to good manufacturing practice (GMP). The presence of heavy metals, pesticides and organic phosphate was controlled by high pressure liquid chromatography (HPLC). Thin layer chromatography, using a TLC-scanner, was used to fingerprint each batch of every individual component (e.g. essential oils). Terpenes were monitored by gas chromatography and mass spectrometry. Batches were rejected when results of fingerprinting differed substantially from that of reference material.
Cell lines, culture conditions and transient transfection procedures
CEM-C7H2, a glucocorticoid-sensitive sub-line of CCRF-CEM-C7 cells [13], was cultured in RPMI 1640 medium (PAA Laboratories, Linz, Austria) supplemented with 10% heat-inactivated fetal calf serum (Biochrom, Berlin, Germany), 100 U/ml penicillin, 100 μg/ml streptomycin (Bio-Whittaker, Walkersville, MD), and 2 mM L-glutamine (Serva, Heidelberg, Germany) at 37°C in a humidified environment of 95% air and 5% CO2. African green monkey kidney fibroblasts (COS-1, 106/100-mm dish) cells were kept at logarithmic growth phase in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum, and 2 mM L-glutamine in a humidified atmosphere containing 5% CO2.
Two h post-transfection, cells were washed with pre-warmed DMEM and kept in RPMI. 24 h before harvesting the cells PADMA 28 extracts or solvent controls were added.
Stock solutions of PADMA 28 and PADMA Basic were prepared in ethanol (EtOH 70%, v/v), methanol (MeOH), and dimethylsulfoxide (DMSO, 1:1 distilled water) and kept at -20°C. Powdered PADMA 28 (400 mg) was vigorously mixed in a total volume of 2 ml of solvent and shaken for 30 min at 180 rpm at 37°C. The mixture was centrifuged at 2,000 × g for 10 min and the supernatants were sterile-filtered (pore size 0.22 μm). Cells were treated with various concentrations of PADMA 28 or PADMA Basic (final concentration range 0.4 mg/ml (1:500) – 1.0 mg/ml (1:200) solvent).
In parallel experiments, effects of pure ethanol, methanol and DMSO (Merck, Darmstadt, Germany) in concentrations comparable to the highest concentration in diluted PADMA 28 extracts were investigated. Viability of cells after application was determined by the trypan blue exclusion method.
Fluorescence Imaging
48 h after transfection, cover slips were washed twice with PBS (140 mM NaCl, 2.7 mM KCl, 4.6 mM Na2HPO4 × 12 H2O, 1.3 mM NaH2PO4 × H2O), and cells were fixed with a 4% solution of p-formaldehyde for 15 min and if necessary, the nuclei were stained by addition of 1 μg/ml DAPI (4',6'-diamidine-2'-phenylindole dihydrochloride, Roche Molecular Biochemicals) diluted in PBS for 10 min. Finally, the cover slips were washed three times with PBS and mounted on slides with Mowiol mounting solution (Calbiochem, Switzerland). Microscopy studies were performed by employing an Olympus BX50 fluorescence microscope and images were taken with the digital camera Microview TE/CCD1317-K/1 (Princeton Scientific Instruments, USA). Imaging analysis was performed by using the Metamorph imaging software (Universal Imaging Corporation).
Flow Cytometry
Detection of apoptotic cells was by FACS analysis of nuclear propidium iodide stained CEM cells. Briefly, 5 × 105 cells were permeabilized and stained with 750 μl propidium iodide (50 μg/ml in 0.1% Triton X-100/0.1% sodium citrate) and subjected to apoptosis analysis in a FACScan (FL-2 channel, Becton Dickinson, San Jose, CA, USA; equipped with an argon laser). Based on propidium iodide staining, cells in the sub-G1 marker window were considered to be apoptotic. Along with nuclear propidium iodide fluorescence, light scattering was also measured. According to the light scattering analysis, apoptotic nuclei are recognized as being smaller (lower forward scatter values) and more granulated (higher sideward scatter values). Cell debris and small particles were excluded from the analysis by forward/sideward scatter criteria as described elsewhere [14].
PARP cleavage
CEM-C7H2 cells were seeded in 100-mm diameter dishes. Logarithmically growing cells were washed once with ice-cold phosphate-buffered saline (PBS) and lysed in 0.5 ml of Triton X-100 lysis buffer (50 mM Tris/HCl, pH 7.3, 50 mM NaCl, 5 mM Na2HPO4 × 12 H2O, 5 mM NaF, 5 mM EDTA, 50 μg/ml of both aprotinin and leupeptin, and 2% Triton X-100) for 20 min on ice. Lysates were clarified by centrifugation at 10.000 × g for 10 min at 4°C after a subsequent sonification procedure. Protein concentrations were determined according to Bradford. After boiling the samples for 5 min, equal amounts of proteins were separated on a 10% SDS-polyacrylamide gel electrophoresis. Transfer to Immobilon-P PVDF membranes (Millipore, Vienna) was in 25 mM Tris/HCl, pH 7.5, 193 mM glycine, 20% (v/v) methanol, and 0.1% SDS at 300 mA constant for 1 h. After blotting, membranes were soaked for 10 min in methanol and washed three times in TBST (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, and 0.5% (v/v) Tween 20). Membranes were blocked overnight with TBST containing 5% (w/v) skimmed milk powder at 4°C and then probed with mouse anti-PARP antibody (diluted at 1:200 in blocking solution, Calbiochem, catalogue number 30-100UG) for 1 h at room temperature. After washing in TBST (4 times for 15 min) the blots were incubated with horseradish peroxidase-conjugated anti-mouse IgG (1:2000 dilution in blocking solution) for 45 min at room temperature. Membranes were washed again in TBST (4 times, 15 min) and immuno-reactive bands were visualized by enhanced chemiluminiscence detection (ECL, Amersham, Vienna, Austria).
Results
Proliferation, inhibition and morphological alterations of CEM-C7H2 cells induced by PADMA 28. CEM-C7H2 cells were treated with various concentrations of PADMA 28. Proliferation was determined by counting cell numbers in a Coulter counter. Figure 1 shows that application of various concentrations of PADMA 28 (0.4 mg/ml; 1.0 mg/ml) inhibited CEM cell proliferation in a concentration- and time-dependent manner. Adding PBS solved extracts at the same concentrations were less effective than alcoholic extracts, suggesting that pharmacologically active ingredients of PADMA 28 are extracted better by unpolar solvents. DMSO alone at the concentrations applied did not have any effect on cell proliferation.
Figure 1 Influence of PADMA 28 on cell proliferation in CEM-C7H2 cells. After a growth period of 24 h, CEM-C7H2 cells were treated with different concentrations of PADMA 28 (0.4 mg/ml (1:500) and 1 mg/ml (1:200). After various application periods, cells were harvested, washed with ice-cold PBS and counted in a Coulter counter to determine cell proliferation rates. Cell dimension gates were set at 8,55 and 29,95 μm. Representative cell numbers are shown. No unspecific effects of the solvent controls were detectable during the incubation period in the cell culture. Data are expressed as cell numbers of untreated controls (± SEM, p < 0.01) of at least three independent experiments done in triplicate. For comparison of grouped data Student's t-test was applied. P-values below 0.05 were considered to indicate significant differences.
In cultivated CEM cells treated with PADMA 28, morphological alterations of the nuclei characteristic for apoptosis were visible. This is shown in panel D of Figure 2.
Figure 2 Morphological alterations of CEM-C7H2 cells after treatment with PADMA 28. CEM-C7H2 cells were either treated with solvent (DMSO) or 0.4 mg/ml of PADMA 28 for 48 h. Cells were stained with DAPI (right panels C/D; 100x) or left unstained (left panels A/B; 10x). Nuclear body formation was examined by fluorescence microscopy. Morphological alterations were analysed by employing transmitted light microscopy (Zeiss) of cells growing on glass cover slips.
PADMA 28-induced subG1 transition
The G1 phase and the G1/S transition state of the cell cycle are critical points at which the cell assesses whether it should enter another full round of division, stop growing, or die.
Transition of cell populations from G1/Go to subG1 phase is a frequent sign of apoptosis. As shown in Figure 3, application of 1 mg/ml PADMA 28 for 48 h resulted in accumulation of approximately 50% of cells in subG1 phase.
Figure 3 PADMA 28-induced subG1 transition. SubG1 transition was detected by employing FACS analysis as described under Methods. Representative subG1 populations calculated from FACS histograms are shown. Data are expressed as the percentage of subG1 cells calculated as the means (± SEM., n = 9, p < 0.01) of three independent experiments done in triplicate. As indicated, identical results were obtained with another solvent.
PARP cleavage kinetic
Activation of pre-existing pro-caspases takes place in a hierarchically ordered sequence starting with activation of initiator caspases and leading to activation of effector caspases such as caspase 3. Caspases 3 cleaves a number of proteins, a prominent substrate being poly (ADP-ribose) polymerase (PARP) [15]. PARP is a zinc-finger DNA-binding enzyme that detects DNA strand breaks generated by genotoxic agents such as oxygen radicals. PARP is activated at an intermediate stage of apoptosis and is inactivated by proteolytic cleavage at a late stage by casapase 3 and caspase 7. We raised the question whether treatment of CEM cells with PADMA 28 causes a caspase 3- or 7-mediated proteolytic cleavage of PARP. Figure 4 shows this was indeed the case. Treatment of CEM-C7H2 cells for 34 h caused a distinct proteolytic fragmentation of PARP, with accumulation of the 89 kDa fragment (lower bands) and concomitant disappearance of the full-size 113 kDa (upper bands) fraction (Figure 4, slot 4, 6, and 8). As was the case with growth inhibitory effect, PBS solutions of PADMA 28 were less effective than ethanolic, methanolic or DMSO extracts.
Figure 4 Kinetics of PARP cleavage. CEM-C7H2 cells were cultivated as described under Methods. Following treatment of cells with 0.1 mg/ml PADMA 28 for 24 h, PARP cleavage was detected by Western blotting. A representative cleavage pattern out of three independent experiments is shown. Upper bands show uncleaved PARP (113 kDa), lower bands correspond to a PARP cleavage product of 89 kDa. As shown for the growth inhibitory effect, PBS solutions were less effective than ethanolic or methanolic ones.
PADMA 28 interferes with Bcl-2 function
Whether a cell lives or dies is largely determined by the Bcl-2 function [11]. If PADMA 28 triggers apoptosis by interfering with pro-survival Bcl-2 signalling, it should be possible to diminish the pro-apoptotic signal by overexpression of Bcl-2. As shown in Figures 5/6, PADMA 28-mediated induction of apoptosis was diminished (Figure 5, bar 6 and 8) by transient overexpression of Bcl-2 in a CEM cell sub-line in which expression of the corresponding trans-gene was induced by withdrawal of doxocycline (Bujard system, CEM C7H2 Bcl-2 TET off).
Figure 5 PADMA 28 interferes with Bcl-2 function. CEM-C7H2 as well as CEM-C7H2Bcl2-TET-off cells were cultivated in the presence (Bcl2 repressed) or absence (Bcl2 expressed) of 200 μg/ml tetracycline; 24 h after seeding, cells were treated with a solution of 0.4 mg/ml PADMA 28 (1:200 EtOH). After 24 and 48 h, cells were harvested and analysed by FACS as described under Methods. Data are expressed as percent of untreated solvent controls (± SEM, ± SEM, p < 0.03) of at least three independent experiments done in triplicate.
Figure 6 Corresponding FACS blots of figure 5.
Bcl-2 expression is controlled, to some extent, by liberation of the Th1-type cytokine IFN-γ [16]. Hence we would like to suggest that PADMA 28 may modulate apoptogenic pathways controlled by interferon-γ.
In this context it should be mentioned that PADMA 28 has been reported to modulate interferon-γ-induced tryptophan degradation and neopterin production in human peripheral blood mononuclear cells (PBMC) from healthy donors [6].
Discussion
Here we show that PADMA 28 added to T cell-derived lymphatic leukaemia cells in culture has a pronounced concentration- and time-dependent anti-proliferative effect. The inhibition of cell proliferation is accompanied by nuclear body formation a morphological signs of apoptosis. Furthermore hallmarks of apoptosis, such as the accumulation of cells in subG1 phase, and fragmentation of poly (ADP-ribose) polymerase (PARP) could be detected.
A growing body of evidence demonstrates that cell survival and cell death are largely determined by the function of Bcl-2 family members. Here we show that induction of apoptosis by PADMA 28 can be effectively counteracted by overexpression of the pro-survival protein Bcl-2. Based on our data we suggest that PADMA 28 may act in a pro-apoptotic fashion by interfering with Bcl-2 signalling.
Bcl-2 expression is, partially, controlled by interferon-γ [16,17]. An influence of PADMA 28 on interferon-γ-mediated signals was recently demonstrated by Neurauter et al. [6] in peripheral blood mononuclear cells (PBMCs). They found that PADMA 28 has a pronounced effect on interferon-γ-stimulated neopterin production and tryptophan degradation in stimulated PBMC [6]. This observation appears to be of particular importance since cytokine-induced tryptophan degradation by indoleamine (2,3)-dioxygenase (IDO) is a potent mechanism to induce apoptosis and T-cell tolerance via deprivation of the essential amino acid tryptophan and by the pro-apoptotic activity of several tryptophan catabolites formed upon this enzyme reaction [18]. Moreover, evidence is accumulating that activation of IDO may represent an immune escape mechanism of tumour cells [19].
Assuming that stimulation of PBMC by mitogens elicits oxidative stress in culture, it seems reasonable to suggest that the above-mentioned effect of PADMA 28 is due to its ability to function as an antioxidant. This assumption is in agreement with findings that PADMA 28 decreases iNOS mRNA levels of mouse macrophages [5], a fundamental process that might contribute to the anti-inflammatory activities demonstrated by others [2-7].
Although PADMA 28 has been demonstrated to inhibit cell proliferation and cause apoptosis under cell culture conditions, one major question is left unanswered: is there any therapeutic window for using a multifactorial herbal remedy, such as PADMA 28, as a supplement in the treatment of leukaemia patients?
Chemotherapeutic agents are most effective when administered in combination (e.g. CHOP, chemotherapeutic multifactorial regimen comprising of doxorubicin, cyclophosphamide, vincristin and prednisone). Traditional Chinese or Tibetan medicine uses mixtures of naturally occurring herbs and herbal extracts for cancer therapy. Many of the commonly used herbal remedies that have been tested have yielded activities that disappeared when the extracts were fractionated into individual chemical components. We agree with Curtis T. Keith et al. (Nature Reviews, Drug Discovery, 2005) [20] that a focus on multicomponent therapeutics might lead to new insights in drug discovery and that multifactorial therapeutics are necessary for the treatment of multifactorial diseases. In agreement with this concept, PADMA 28, a multifactorial herbal formula, might be similarly effective as a chemotherapeutic agent administered as a supplement, and/or be able to minimize the toxic side effects of such a regimen.
It is not clear to what extent one can extrapolate results from our cell culture experiments to possible effects in humans. Not all of the compounds can be efficiently absorbed, and a gradient exists between the concentrations present in the extracts and the ones established in the blood. Nevertheless, one can assume the presence of all ingredients in the gastrointestinal tract at reasonable concentrations, which makes anti-inflammatory and pro-apoptotic potency more likely. This assumption is further reinforced by the fact that PADMA 28 is successfully used to treat intermittent claudication, atherosclerosis, scleroderma, multiple sclerosis, and chronic hepatitis.
Clinical trials are required for determining whether PADMA 28 can be effectively used in the treatment of cancers. However, biochemical studies might help bridge the gap between conventional and alternative tumor therapy strategies. Additional studies focusing on the biological significance of PADMA 28-enhanced apoptosis in various cell systems will greatly extend our understanding of the mode of action of this remedy. In order to clarify the biological function of PADMA 28 in molecular fashion, gene expression profiling experiments by employing Affymetrix technology are currently under way.
Acknowledgements
We are indebted to Stefan Geley for providing us with acute lymphocytic leukaemia sub-cell lines (CEM & CEM-C7H2 & TET-off). Thanks are also due to Rajam Csordas for critical reading of the manuscript and editorial assistance. This work was supported in part by a grant by the Austrian Research Council Grant (FWF), Science Research Centre SFB F021, Cell proliferation and cell death in tumours – Molecular mechanisms underlying the interplay of proliferation and apoptosis.
==== Refs
Drabaek H Mehlsen J Himmelstrup H Winther K A botanical compound, PADMA-28, increases walking distance in stable intermittent claudication Angiology 1993 44 863 7 8239057
Gieldanowski J Dutkiewicz T Samochowiec L Wojcicki J PADMA-28 modifies immunological functions in experimental atherosclerosis in rabbits Arch Immunol Ther Exp 1992 40 291 295
Gladysz A Juszczyk J Brzosko WJ Influence of PADMA 28 on Patients with Chronic Active Hepatitis B Phytotherapy Research 1993 7 44 47
Suter M Richter C Anti- and pro-oxidative properties of PADMA 28, a Tibetan herbal formulation Redox Rep 2000 5 17 22 10905539
Moeslinger T Friedl R Volf I Brunner M Koller E Spieckermann PG Inhibition of inducible nitric oxide synthesis by the herbal preparation PADMA 28 in macrophage cell line Can J Physiol Pharmacol 2000 78 861 866 11100933 10.1139/cjpp-78-11-861
Neurauter G Wirleitner B Schennach KH Ueberall F Fuchs D PADMA 28 modulates interferon-γ-induced tryptophan degradation and neopterin production in human PBMC in vitro International Immunopharmacology 2004 4 833 839 15135323 10.1016/j.intimp.2004.03.010
Ginsburg I Sadovnik M Sallon S Milo-Goldzweig I Mechoulam R Breuer A Gibbs D Varani J Roberts S Cleator E Singh N PADMA-28, a traditional Tibetan herbal preparation inhibits the respiratory burst in human neutrophils, the killing of epithelial cells by mixtures of oxidants and pro-inflammatory agonists and peroxidation of lipids Inflammapharmacology 1999 7 47 62
Costantini P Jacotot E Decaudin D Kroemer G Mitochondrion as a novel target of anticancer chemotherapy J Natl Cancer Inst 2000 92 1042 1053 10880547 10.1093/jnci/92.13.1042
Green DR Evan GI A matter of life and death Cancer Cell 2002 1 19 30 12086884 10.1016/S1535-6108(02)00024-7
Makin G Dive C Apoptosis and cancer chemotherapy Trends in Cell Biol 2001 11 22 27 11146295 10.1016/S0962-8924(00)01865-1
Cory S Huang DCS Adams JM The Bcl-2 family: roles in cell survival and oncogenesis Oncogenes 2003 22 8590 8607 10.1038/sj.onc.1207102
Geley S Hartmann BL Kofler R Ceramides induce a form of apoptosis in human acute lymphoblastic leukaemia cells that is inhibited by Bcl-1, but not by CrmA FEBS Lett 1997 400 15 18 9000505 10.1016/S0014-5793(96)01284-7
Strasser-Wozak EMC Hattmannstorfer R Hala M Hartmann BL Fiegl M Geley S Kofler R Splice site mutation in the glucocorticoid receptor gene causes resistance to glucocorticoid-induced apoptosis in human acute leukaemia cell line Cancer Res 1995 55 348 353 7812967
Sgonc R Wick G Methods for detection of apoptosis Int Arch Allergy Immunol 1994 105 327 332 7981599
Decker P Muller S Modulating poly (ADP-ribose) polymerase activity: potential for the prevention and therapy of pathogenic situations involving DNA damage and oxidative stress Curr Pharm Biotechnol 2002 3 275 283 12164482 10.2174/1389201023378265
Molostvov G Morris A Rose P Basu S Modulation of Bcl-2 family proteins in primary endothelial cells during apoptosis Pathophysiol Haemost Thromb 2002 32 85 91 12214154 10.1159/000065081
Kamei T Inui M Nakamura S Okumura K Goto A Tagawa T Interferon-γ and anti-Fas antibody-induced apoptosis in human melanoma cell lines and its relationship to Bcl-2 cleavage and Bak expression Melanoma Res 2003 13 153 159 12690298 10.1097/00008390-200304000-00007
Fallarino F Grohmann U Vacca C Bianchi R Orabona C Spreca A Fioretti MC Puccetti P T cell apoptosis by tryptophan catabolism Cell Death Differ 2002 9 1069 77 12232795 10.1038/sj.cdd.4401073
Uyttenhove C Pilotte L Theate I Stroobant V Colau D Parmentier N Boon T Van den Eynde BJ Evidence for a tumoral immune resistance mechanism based on tryptophan degradation by indoleamine 2,3-dioxygenase Nature Medicine 2003 9 1269 1274 14502282 10.1038/nm934
Keith CT Borisy AA Stockwell BR Multicomponent therapeutics for networked systems Nature Reviews Drug Discovery 2005 4 1 8 10.1038/nrd1609
|
16138918
|
PMC1232859
|
CC BY
|
2021-01-04 16:39:20
|
no
|
J Carcinog. 2005 Sep 2; 4:15
|
utf-8
|
J Carcinog
| 2,005 |
10.1186/1477-3163-4-15
|
oa_comm
|
==== Front
Chiropr OsteopatChiropractic & Osteopathy1746-1340BioMed Central London 1746-1340-13-111602662510.1186/1746-1340-13-11ReviewAnatomic and functional leg-length inequality: A review and recommendation for clinical decision-making. Part I, anatomic leg-length inequality: prevalence, magnitude, effects and clinical significance Knutson Gary A [email protected] 840 W. 17th, Suite 5 Bloomington, IN, 47404, USA2005 20 7 2005 13 11 11 31 5 2005 20 7 2005 Copyright © 2005 Knutson; licensee BioMed Central Ltd.2005Knutson; 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
Leg-length inequality is most often divided into two groups: anatomic and functional. Part I of this review analyses data collected on anatomic leg-length inequality relative to prevalence, magnitude, effects and clinical significance. Part II examines the functional "short leg" including anatomic-functional relationships, and provides an outline for clinical decision-making.
Methods
Online database – Medline, CINAHL and MANTIS – and library searches for the time frame of 1970–2005 were done using the term "leg-length inequality".
Results and Discussion
Using data on leg-length inequality obtained by accurate and reliable x-ray methods, the prevalence of anatomic inequality was found to be 90%, the mean magnitude of anatomic inequality was 5.2 mm (SD 4.1). The evidence suggests that, for most people, anatomic leg-length inequality does not appear to be clinically significant until the magnitude reaches ~ 20 mm (~3/4").
Conclusion
Anatomic leg-length inequality is near universal, but the average magnitude is small and not likely to be clinically significant.
Leg-length inequalityanatomicback painchiropractic
==== Body
Review
Leg-length inequality (LLI) is a topic that seemingly has been exhaustively examined; yet much is left to be understood. Reviews by Mannello [1] and Gurney [2] on leg-length inequality and Cooperstein and Lisi on pelvic torsion [3] are highly recommended as sources to provide expanded and longer time-frame background information on this topic. The information provided by these authors, however extensive, is incomplete relative to clinical decision-making. Further, several questions have remained largely unanswered regarding anatomic leg-length inequality and the so-called functional short leg, or more accurately, unloaded leg-length alignment asymmetry (LLAA). These include: how common is anatomic LLI, what is the average amount of anatomic LLI, what are the effects of anatomic LLI, how much anatomic LLI is necessary to be clinically significant, and what are the incidental and functional relationships of anatomic LLI to unloaded leg-length alignment asymmetry? The purpose of this review is to highlight current research to answer these questions and help in clinical decision-making.
Methods
In the 1970's studies began to show that clinical measurements of LLI were inaccurate and the use of x-ray, controlling for magnification and distortion, was necessary [4-6]. By 1980 the accuracy of the measurements with the standing x-ray had been established, with Friberg then demonstrating reliability of the method on subjects [7]. For these reasons, this review starts in the 1970's with studies that used the reliable x-ray procedure as described by Friberg.
To answer the question regarding the prevalence of anatomic leg-length inequality, Medline, CINAHL, MANTIS and library searches (using key words "leg-length inequalty") were performed for studies done from 1970–2005. Studies which did not describe, or use the reliably precise radiographic method, or that did not provide their LLI measurement data, were excluded.
Prevalence of anatomic leg-length inequality
Several studies using the precise radiographic method (Table 1) contained data, which quantified LLI in incremental millimetric measurements [8-15]. These studies were combined giving a population of n = 573, with a LLI range of 0–20 mm. The mean LLI was 5.21 mm (SD 4.1 mm) or approximately 3/16". The results of these studies are shown in Figure 1. Six of the studies, with combined population of n = 272, broke their data down into right or left LLI [8-12,14]. Figure 2 shows those results; note the curve is shifted slightly towards leg-length discrepancy on the right. This finding – that the right leg is anatomically shorter more often – is consistent with other studies that have found the left leg to be anatomically longer 53–75% of the time [6,7,9]. Using the same studies [8-12,14] to compare the magnitude of the discrepancy of right (n = 140) and left (n = 114) legs finds only a 0.84 mm difference, which is not statistically significant (p = 0.08, t-test). This means that while the right leg is anatomically short more often, the amount of the discrepancy is no greater than a short left leg.
Table 1 Studies using reliable means of determining magnitude of anatomic leg-length inequality
Study Population "N" (573) Subjects/Notes Controls Av LLI (SD)
Gross R. 1983 Male marathon runners, age 24–49 33 No deleterious effect of the LLI 4.9 mm (3.8)
Venn et al 1983 Randomly chosen patients 60 5.4 mm (4.0)
Cleveland et al 1988 Low back pain patients 10 Standing and supine x-ray 4.7 mm (5.8)
Hoikka et al 1989 Chronic low back pain patients 100 4.9 mm (3.6)
Beattie et al 1990 Clinical subjects, age 22–60 19 10 with history of LLI or lower extremity or back pain 9 healthy 6.8 mm (5.7)
Soukka et al 1991 Four defined occupational and gender groups, age 35–54 247 194 with prior back pain (>12 mo ago and during last 12 mo with and without disability) 53 who never had back pain 5.0 mm (3.9)
Rhodes et al 1995 New LBP patients Chiropractic practice 50 Age 18–40 26 men 24 women 6.3 mm (4.1)
Mincer et al 1997 Volunteers 54 no history of back pain in last 6 months 10 men 44 women 2.4 mm (1.8)
Figure 1 "Incidence" of anatomic leg-length inequality magnitude.
Figure 2 Magnitude of anatomic leg-length inequality; right vs. left.
Four of the radiographic studies [8,10,12,15] identified measured LLI subjects by gender (n = 116). There was no difference (p = 0.87, t-test) between male and female LLI as shown in Table 2, suggesting that gender plays little role in the amount of anatomic LLI. One study [12] provided data on subject height (n = 19), which was plotted against LLI giving only a fair correlation coefficient of 0.31. However, Soukka et al, using a much larger number of subjects (n = 247) did find a correlation between height and LLI (p = 0.02)[13]. Men, being taller than women on average, would be expected to show a larger LLI, but did not. The discrepancy in these data is difficult to explain.
Table 2 Relationship between gender and anatomic leg-length inequality
LLI and Gender (Refs 8,10,12,15) N Mean LLI (mm)
Male 58 5.1 (4.3)
Female 58 5.2 (4.6)
P = 0.87 (t-test)
Seven of the studies identified subjects with LLI as being symptomatic (n = 347) or asymptomatic (n = 165) [8-10,12-15]. Symptoms included a variety of kinetic chain (knee, hip) problems and low back pain. Asymptomatic was variously defined from no complaints, to no back pain in the last six months [15], to no low back pain in the last 12 months [13]. Symptomatic subjects had a mean LLI of 5.1 mm (SD 3.9); asymptomatic subjects had a mean LLI of 5.2 mm (SD 4.2). There is no statistical difference in the LLI between these two groups (p = 0.75, t-test). The mean LLI for these groups is virtually identical to the overall combined mean, suggesting that the average LLI is not correlated to symptomatic problems, especially low back pain.
Recognizing that measurements to the precision of a millimeter will be prone to error, other studies – again, using precise radiographic methods – have examined LLI within a measured range [7,16,17]. These findings, combined with the millimetric measure studies, are noted in Figure 3, and provide an even larger pool of data for LLI. This data table shows, for example, that in a pooled population of 2,978 people, 20.1% had a LLI of 10 mm or more. Collecting x-ray data from 421 subjects with low back pain from an osteopathic manipulative practice, Juhl et al [18] reported on the incidence of leg-length and sacral base unleveling. The data from Juhl et al indicated that 43% of those examined had LLI of 10 mm or more, twice the rate noted from the pooled data in this review. A significant difference of Juhl et al's methods of examination was that the central ray was directed at the level of the sacral base, and not the femoral heads. Due to this methodological difference, lack of reported reliability of this method, and the significant disagreement with others as to incidence, the data from Juhl et al regarding the incidence of anatomic leg-length inequality was not used.
Figure 3 Ranges of anatomic leg-length inequality.
Using the data from the millimetric measurement, 90% of the population has some anatomic leg-length asymmetry. This finding is in accord with other studies [19,20]. Larger LLI – more than 20 mm (~ 3/4") – was calculated in a population of 2.68 million, to be 1 in 1000 [21]. References will be made later in this paper to the data compiled in these two tables.
Finally, in a retrospective study of 106 consecutive patients, Specht and De Boer report on the use of 14" × 36" x-ray films to determine LLI [22]. This x-ray method, less reviewed than the methods noted above, does not direct the central ray at the femoral heads and therefore uses a mathematical formula to take the effect innominate rotation into account in measuring LLI. The results calculated from the data presented showed an average LLI of 5.5 mm (SD 3.9), which is nearly identical to the multi-study average noted above.
Effects of LLI
The most common effect of anatomic LLI is rotation of the pelvis and/or innominate bones – often referred to as pelvic torsion – in the sagittal and/or frontal planes [3,23-25]. Mechanically, in the standing position, the weight of the body in the pelvis induces a force vector through the hip joints and towards the feet. With asymmetry of the leg-lengths, the pelvis, being pushed down on the femoral heads, must rotate or torsion. The innominate movement tends to be anterior on the side of the anatomically short leg and posterior contralaterally [23,26]. In studies of pelvic rotation imposed by foot lifts, there was an approximately linear relationship in pelvic torsion as the leg was lengthened from 1/4 to 7/8" [23]. A chart, based on the work of Cummings et al, shows the degrees of torsion relative to lengthening of the left leg (Figure 4). Note that the artificial lengthening of the left leg caused more rotation of the contralateral hemipelvis in an anterior direction – the short leg side – than posterior rotation ipsilaterally.
Figure 4 Pelvic rotation with left heel lift (Cummings).
The relationship of LLI to pelvic torsion is supported by the data of others [27]. Walsh et al [24] found that pelvic obliquity was the most common method of compensating for LLI up to 22 mm. With larger amounts of leg-length inequality, subjects begin to develop flexion of the knee in the long leg [24]. While the degree of pelvic torsion due to the imposition of lifts tends to be linear, there are many factors – including innominate asymmetry, freedom of SI joint movement, and hypertonic suprapelvic muscles – that can affect pelvic torsion. Several authors emphasize that it is a mistake to assume that the side and amount of LLI can be reliably deduced from pelvic crest unleveling [17,26,28].
Other effects of LLI and pelvic torsion have been demonstrated by Giles et al [29,30]. These compensations include alterations and asymmetry of lumbosacral facet joint angles, postural scoliosis, concavities in the vertebral body end-plates, wedging of the 5 th lumbar vertebra and traction spurs. However, no relationship of these findings to symptoms was claimed.
Along the lines of symptomatic problems associated with LLI compensations, Levangie attempted to quantify pelvic asymmetry in a loaded (standing) position without x-ray by using precise location of anatomic landmarks [31]. The objective was to see if pelvic torsion – the most common compensation for LLI – was correlated with back pain. It was not. In another study, a pelvic level – a device with a weighted gravity line superimposed on a scale in one-degree increments clamped in place on the palpated superior aspects of the iliac crests – was used to examine a group of non-clinical subjects [32]. There was no correlation of self-reported back pain, frequency or severity, to pelvic unleveling. However in those subjects with measurable pelvic unleveling (29 of 64 subjects), 61% had a high left iliac crest, which may be evidence of the greater incidence of a longer left leg [32]. A final study, using radiography to determine pelvic obliquity, examined subjects with (n = 93) and without (n = 76) chronic low back pain (defined as low back pain of at least 3 months) [33]. This study found no difference in the pelvic obliquity between subjects with and without chronic back pain, obliquity was prevalent and equally distributed in both groups.
These studies examining pelvic obliquity indicate that this type of postural distortion, be it from LLI or bony asymmetry, is not related to back pain, and does not seem to be clinically significant. The next, more difficult and controversial question is, what is the clinical significance of LLI, and at what magnitude?
How much anatomic LLI is clinically significant?
Mannello remarked that the clinical significance of LLI was "perhaps...dependent on several factors, including the degree of inequality, the ability of the pelvis and spine to compensate for the inequality and associated conditions or problems" [1]. While this statement is undoubtedly true, this paper will attempt to quantify what ranges of anatomic LLI are clinically significant, that is, being associated with back pain, injury, muscle strength asymmetry or other physiologic changes. Unless noted, all the studies reviewed here have been selected because they used the more accurate radiological methods to determine anatomic LLI.
When one examines references alluding to the clinical significance of anatomic LLI, Friberg's 1983 study [7] is most often cited. Friberg collected data on 1,157 subjects; 798 with chronic LBP and a control group of 359 with no LBP. The data Friberg collected on the prevalence of LLI in a normal population is very similar to that found in the compilation outlined in this paper. The prevalence of LLI 10 mm or greater was 15.6%. This review found the figure to be 14.8%; Friberg showed the incidence of LLI 15 mm or greater at 2.2%, this review 2.6%. Unlike the population compiled in this review however, Friberg's data were obtained from patients at a military hospital and represent a high percentage of subjects exposed to extreme and repetitive loading.
Friberg also reported "LLI was 5 mm or more in 75.4% of the patients with LBP and 43.5% of the controls. The difference is statistically significant (P < 0.001) using a Chi-squared test" [7]. Anatomic LLI greater than 20 mm was previously shown to be the putative limit for spontaneous compensation of the pelvis to postural asymmetry. If these subjects are eliminated from Friberg's data, the association of anatomic LLI with LBP drops somewhat.
Chronic low back pain (CLBP) affects about 21% of the population [34,35]. One would expect this percentage to be higher if, as Friberg found, that 5 mm of LLI is a causative factor, given that 50% of the population has LLI of 5.2 mm or greater. Figure 5 shows the relative "incidence" of chronic low back pain to LLI using Friberg's data. As can be seen, Friberg's putative correlation really becomes demonstrable when LLI is above 15 mm, at 5.3 times the prevalence of CLBP.
Figure 5 "Incidence" of chronic low back pain with anatomic leg-length inequality (Friberg).
In defending the results and their interpretation in a letter-to-the-editor, Friberg wrote, "... I have always pointed out that LLI of less than 5 mm has no relationship with lumbar scoliosis or back pain. I have also emphasized that even marked LLI per se [emphasis in original] neither produces LBP nor contributes to its development if a person is not continually exposed to prolonged standing or gait, e.g., during daily work, military training, and sporting activities" [36]. So, Friberg notes that relatively small amounts of anatomic LLI may only be clinically significant relative to certain conditions such as prolonged and/or repetitive loading – which describes the population in Friberg's study – and not as a generality, as the study is often referenced to support.
Friberg's data represents the low end of anatomic LLI that is hypothesized to be clinically significant. At the high end, in a review of the biomechanical implications of leg-length inequality, others write that LLI less than 30 mm is mild and the clinical significance questionable [25,37]. This large range – from 5 mm to 30 mm – is the likely reason behind the lack of consensus as to the clinical significance of LLI. The answer presumably lies somewhere in between.
Giles and Taylor [30] reported that LLI of 10 mm or larger was found significantly more often in a group with chronic low back pain. No data was given as to the mean LLI or the distribution in the CLBP group, only that the LLI was greater than 9 mm. They found LLI of 10 mm or more in 18% of the CLBP population (n = 1309), and only 8% of the normal population (n = 50). The pooled data (n = 164) of asymptomatic subjects in this review [8,10,12-15] finds 15.5% of this population with LLI of 10 mm or greater. The data compiled from all pooled studies – both symptomatic and symptom free – shows LLI of 10 mm or greater in 15% of the population. These results raise questions about whether the prevalence of LLI found in Giles and Taylor's normal population is representative and whether CLBP is indeed related to LLI in the 10 mm range.
Similarly, Kujala et al found athletes with patellar apicitis (jumpers knee) had a significantly larger LLI (5.8 mm, SD 4.5) than an asymptomatic control group (3.0 mm, SD 2.3) [38]. The mean LLI in the Kujala et al control group (n = 20) is significantly less than the pooled asymptomatic subjects (n = 164) in this review (5.2 mm, SD 4.2) [8,10,12-15] and may be related to the smaller sample size, or the unique group sampled ("healthy" athletes). Regardless, the Kujala et al control group does not represent the asymptomatic general population based on the evidence examined in this review paper. Kujala et al also studied military conscripts (n = 32) who developed knee pain during their initial 8-weeks of training and compared them to a group that did not develop knee pain (n = 28). Those who had knee pain had a significantly larger LLI; 8.0 mm (SD 5.9) versus 4.1 mm (SD 2.9) at p = 0.003 (t-test) [39]. While the magnitude of LLI in the control group in this study is much closer to the normal demonstrated in this review paper (5.2 mm versus 4.1 mm), the magnitude of the control group LLI is also very close to that of the patellar apicitis group in the athlete study. One might question why athletes are more likely to develop knee pain with an average LLI of 5.8 mm, but training soldiers with an average LLI of 4.1 mm are not? Further, there was no correlation between the injured knee and the side of the short leg, which would be expected if the short leg were the predisposing factor.
In a survey of 247 working age men and women looking for the presence of LLI, Soukka et al [13] examined and compared statistically matched groups with and without LBP. Their results showed no increased risk of back pain with a LLI of 10–20 mm, and the relationship between LLI of more than 20 mm and back pain were not conclusive. These results differ markedly from that of Friberg, prompting the letter-to-the-editor noted above [36]. In the exchange between Friberg and Soukka et al, both agree that the significance of LLI may depend more on prolonged and repetitive loading, a common sense idea previously expressed by Subotnick [40].
One of the areas of research into the clinical significance of LLI has been in relation to femoral fracture and total hip replacement surgery. Gibson et al found that in 15 patients, at least 10 years after shortening due to femoral fracture (average 3 cm, range 1.5 – 5.5 cm), there was no significant discomfort, structural abnormalities or degenerative changes as a result of the leg length discrepancy [41]. Edeen et al followed 68 patients with a mean LLI of 9.7 mm for an average of 6.6 years after hip replacement surgery [42]. They were not able to demonstrate a relationship between LLI and low back pain. Another study of 200 post total hip replacement surgeries used validated functional outcome scores (Harris hip score and the SF-36 Health Survey) to examine the relationship of imposed LLI to functional outcome [43]. This study found that leg lengthening (up to 35 mm) or shortening (up to 21 mm) did not correlate with decreased function, comfort or satisfaction six months after the operation. A retrospective study of 6,954 total hip arthroplasty patients over a 7 year period found only 21 (0.3%) had symptoms related to post-surgery leg length inequality symptomatically severe enough (primarily back and hip pain) to require a second surgery to equalize leg length [44]. The mean LLI of the patient's who received revision arthroplasty was 3.6 cm (± 1.2 cm, range 2.0 cm to 7.0 cm). The results of these studies of hip replacement are somewhat surprising given that the LLI was induced at an older age when the ability of the pelvis, SI joints and soft tissues to compensate for this asymmetry would likely be reduced.
In examining the effects of LLI from childhood, Yrjönen et al [45] did a follow-up study of 81 patients with Perthes' disease and a mean LLI of 12 mm. The follow-up time was an average of 35 years (range 28–47). They found that most of the patients had no back pain, and concluded that back pain was not a significant problem after Perthes' in spite of frequent LLI. Another study of adults (mean age 28) with large LLI since childhood – mean 29.1 mm – found no complaints of back pain or degenerative changes. Lumbar scoliosis was minor in those with LLI of less than 22 mm [46].
In most of these studies, follow-up was years to decades, and LLI means from ~ 10 mm to 30 mm, yet none could demonstrate a significant correlation to back pain. Given these findings, the average 5 mm anatomic leg-length differential does not appear to be significant, even with prolonged and repetitive loading. Based on these studies, childhood-onset LLI up to at least 20 mm (~ 3/4") does not seem to be clinically significant.
Another category where LLI can cause sudden, abnormal loading of the lumbar-pelvic structure is in athletic and military training. Gross [10] examined LLI in a group of marathon runners. He found that leg length discrepancy less than 25 mm did not appear to have a deleterious effect. In a study of stress fracture and LLI in Finnish army conscripts, Friberg [47] found those with LLI ≥ 10 mm had stress fractures 10% more frequently than healthy (no known stress fracture) controls. No statistical analysis was described, so it is not known whether the increase in stress fracture incidence from 20.1% (controls) to 30% (patients) is significant. Friberg did find that in parachutists, those with LLI 10 mm or more (15.7% of 102, n = 16), 50% had stress fractures. This does point to an association between LLI of over 10 mm, extreme loading and stress fracture, however, the small "n" of 16 did not allow for a statistical analysis. In a study of athletes (n = 46) for anatomic LLI as a risk factor in stress fractures, Korpelainen et al [48] found the mean LLI of the patient group to be 4.9 mm. While sympathetic to the possibility of the association between LLI and stress fracture, they found no relationship.
Again, the average amount of LLI (5 mm) does not appear to be clinically significant with substantially increased and repetitive loading. Only when the increased loading is abrupt and severe (Friberg's parachutists) is a strong correlation established between LLI of 10 mm and a pathologic condition (stress fracture). Given the findings in these studies, LLI below 10 mm, even with heavier repetitive loading, does not appear to be clinically significant. LLI between 10 – 20 mm increases the chances of clinical significance, but outside of severe, abrupt loading, the evidence is lacking. Based on these studies, it would appear that childhood-onset LLI of up to 15 – 20 mm does not seem to be clinically significant.
The effect of LLI on physiological function has also been explored, and can shed some light on a possible range of clinical significance. It has been presumed that anatomic LLI, because of its effects on structure, causes muscular hypertonicity and changes in strength and/or coordination [28]. Mincer et al [15] expected LLI, (because of presumed stressful mechanical effects on the lumbar spine by virtue of the asymmetrical loading) would cause earlier and greater fatigue of trunk muscles, and tested that hypothesis. The average inequality in the LLI group (n = 18) was 10 mm. They found no difference between the LLI and no LLI groups relative to muscle fatigue or neuromuscular control. Yen et al [49] examined muscular performance on trunk extension in a group of young men with estimated LLI of 10 – 15 mm, both with and without a lift used to equalize LLI. There was no statistically significant effect of the lift and equalization of LLI on any of the variables tested. Murrell et al [50] examined standing balance in subjects with LLI of at least 9.5 mm versus those with no LLI and found no difference. They concluded that individuals with anatomic LLI are not less stable than those without during quiet stance, and that the probable reason for this finding is long-term adaptation by the neuromuscular system to the LLI.
The last two studies [49,50] relied on more inaccurate measures to determine LLI, so the results are suspect. However, in these studies, care was taken to classify and examine only those with far end-range amounts of asymmetry as having LLI; in all studies that amount was over 10 mm.
In a study of LLI and analysis of gait, Goel et al found no significant differences in joint movement with the imposition of a 1.25 cm leg length differential via a shoe (not just heel) lift [51]. Based on their findings, they suggest that, "...the body is well able to compensate for minor LLD [leg length differential] of up to 2 cm. Correction of an LLD of this magnitude for biomechanical reasons alone does not appear indicated". Another study of gait with LLI imposed via foot lifts found that a 2.3 cm lift produced no changes in gait or hip forces and moments [52]. A study of subjects with pre-existing LLI found that a mean LLI of 2.5 cm was necessary to produce an asymmetrical gait [53]. A study of the effect of LLD in children (n = 20, 9.0 ± 3.9 years) on found gait asymmetries only with LLD >2.0 cm [54]. White et al found LLD between 1 – 3 cm, whether simulated or real, resulted in unequal loading of limbs when walking, and recommended considering shoe lifts to equalize leg lengths [55]. Finally, in examining the effects of imposed foot lifts on oxygen requirements, one study found no statistical difference even with a lift of 3 cm during running [37]. Another found it was necessary to impose a LLI of between 2 – 3 cm in older adults in order to cause increased oxygen consumption and perceived exertion [56].
As in the previous groups (general working population, long-term loading, and heavy loading) the effects of LLI in the order of 10 mm relative to muscle strength, coordination and gait and oxygen consumption do not appear clinically significant. The evidence in these studies is less compelling because of the measurement methods, the concentration of testing around the 10 mm mark and imposition of LLI, which does not give the body time to compensate. However, there is no reason to believe that those physiological measures are any more sensitive to LLI than the other measures noted previously.
These findings – that LLI in the range of 20 mm (~ 3/4"), regardless of prolonged or repetitive loading, does not result in back pain or other clinically significant symptom, seems to preclude the need for heel lifts in most cases. However, there will always be individual exceptions, and there may be some general exceptions.
Gofton and Trueman found a strong association between leg length and unilateral osteoarthritis (OA) in the supero-lateral region of the hip on the side of the anatomically long leg [4]. In their study, all subjects with this type of OA "had led healthy active lives prior to the onset of hip pain", and few subjects were aware of any difference in leg length. The authors point out that this form of OA has its onset around the age of 53. While concluding that anatomic LLI in the order of 1/2" to 1" (13 – 25 mm) is associated with the development of a unilateral OA hip, Gofton and Trueman acknowledge that many with this anatomic asymmetry fail to develop this condition, suggesting that factors other than the disparity are also important. An important area of investigation would be to determine these other factors to provide a clearer picture of who may be at risk.
Further data suggesting exceptions to the conclusions drawn above regarding the effects of mild anatomic LLI come from Triano [57]. He demonstrated balancing of asymmetric electromyographic paraspinal muscle activity in 51% of subjects with low back pain by using an average heel lift of 22 mm. These results indicate that changes in leg length of ~3/4" or greater results in active – muscular – compensation which, if prolonged, may become painful. Bringing the pelvis back towards a neutral orientation and decreasing active muscular compensation may explain why the use of heel lifts under the short leg appears to be an effective treatment in some complaints of back pain [7,16,58]. To explain these results, the functional "short leg" will be examined in Part II.
In summary, childhood-onset anatomic leg-length inequality appears to have little clinical significance up to 20 mm. Several authors agree [2,25,59], most recently with Kakushima et al who stated: "Therefore, although conflicts in the literature exist, 3 cm of LLD [leg length discrepancy] can be characterized as a minimum LLD, which should be treated in the clinical practice" [60]. This estimation of clinical significance dovetails nicely with the findings on the effects of LLI, particularly pelvic torsion [23]. Passive structural changes – pelvic torsion, mild lumbar scoliosis, facet angulation, changes in muscle length – seem capable of compensating for anatomic LLI of up to 20 mm. Past the ~ 20 mm point, passive structural changes give way to active muscular compensatory measures.
Conclusion
The purpose of this paper was to review radiographic research regarding anatomic leg-length inequality; prevalence, mean magnitude, effects, clinical significance and relationship to unloaded leg-length alignment asymmetry. Ninety per cent of the population has some anatomic leg-length inequality; the average was found to be 5.2 mm. Based on the research reviewed, childhood anatomic LLI of less than 15 mm in situations of repeated and/or heavy loading, or less than 20 mm (~ 3/4") under normal conditions, is not likely to cause symptoms requiring treatment.
Competing interests
The author(s) declare that they have no competing interests.
==== Refs
Mannello DM Leg Length Inequality J Manipulative Physiol Ther 1992 15 576 590 1469342
Gurney B Leg length discrepancy Gait Posture 2002 15 195 206 11869914 10.1016/S0966-6362(01)00148-5
Cooperstein R Lisi A Pelvic torsion: anatomic considerations, construct validity, and chiropractic examination procedures Top Clin Chiro 2000 7 38 49
Gofton JP Trueman GE Studies in osteoarthritis of the hip: Part II. Osteoarthritis of the hip and leg-length disparity CMA Journal 1971 104 791 799
Clarke GR Unequal leg length: an accurate method of detection and some clinical results Rheum Phys Med 1972 11 385 390
Fisk JW Baigent ML Clinical and radiological assessment of leg length NZ Med J 1975 477 480
Friberg O Clinical symptoms and biomechanics of lumbar spine and hip joint in leg length inequality Spine 1983 8 643 651 6228021
Cleveland RH Kushner DC Ogden MC Herman TE Kermond W Correia JA Determination of leg length discrepance. A comparison of weight-bearing and supine imaging Invest Radiol 1988 23 301 4 3372194
Hoikka V Ylikoski M Tallroth K Leg-length inequality has poor correlation with lumbar scoliosis Arch Orthop Trauma Surg 1989 108 173 75 2525016 10.1007/BF00934262
Gross RH Leg length discrepancy in marathon runners Am J Sports Med 1983 11 121 124 6869652
Venn EK Wakefield KA Thompson PR A comparative study of leg-length checks Eur J Chiropractic 1983 31 68 80
Beattie P Isaacson K Riddle DL Rothstein JM Validity of derived measurements of leg-length differences obtained by use of a tape measure Phys Ther 1990 70 150 7 2304973
Soukka A Alaranta H Tallroth K Heliovaara M Leg-length inequality in people of working age. The association between mild inequality and low-back pain is questionable Spine 1991 16 429 431 1828627
Rhodes DW Mansfield ER Bishop PA Smith JF The validity of the prone leg check as an estimate of standing leg length inequality measured by x-ray J Manipulative Physiol Ther 1995 18 343 346 7595107
Mincer AE Cummings GS Andrew PD Rau JL Effect of leg length discrepancy on trunk muscle fatigue and unintended trunk movement J Phys Ther Sci 1997 9 1 6 10.1589/jpts.9.1
Giles LGF Taylor JR Low-back pain associated with leg-length inequality Spine 1981 6 510 521 6458101
Gross MT Burns CB Chapman SW Hudson CJ Curtis HS Lehmann JR Renner JB Reliability and validity of rigid lift and pelvic leveling device method in assessing functional leg length inequality JOSPT 1998 27 285 294 9549712
Juhl JH Cremin TM Russell G Prevalence of frontal plane pelvic postural asymmetry – part 1 J Am Osteopath Assoc 2004 104 411 21 15537798
Lawrence D Lateralization of weight in the presence of structural short leg: a preliminary report J Manipulative Physiol Ther 1984 7 105 108 6747485
Korpelainen R Orava S Karpakka J Siira P Hulkko A Risk factors for recurrent stress fractures in athletes Am J Sports Med 2001 29 304 10 11394600
Guichet J-M Spivak JM Trouilloud P Grammont PM Lower limb-length discrepancy. An epidemiological study Clin Orthop Rel Res 1991 272 235 241
Specht DL De Boer KF Anatomical leg length inequality, scoliosis and lordotic curve in unselected clinical patients J Manipulative Physiol Ther 1991 14 368 375 1919374
Cummings G Scholz JP Barnes K The effect of imposed leg length difference on pelvic bone symmetry Spine 1993 18 368 373 8475440
Walsh M Connolly P Jenkinson A O'Brien T Leg length discrepancy – an experimental study of compensatory changes in three dimensions using gait analysis Gait Posture 2000 12 156 61 10998613 10.1016/S0966-6362(00)00067-9
McCaw ST Bates BT Biomechanical implications of mild leg length inequality Br J Sp Med 1991 25 10 13
Young RS Andrew PD Cummings GS Effect of simulating leg length inequality on pelvic torsion and trunk mobility Gait Posture 2000 11 217 23 10802434 10.1016/S0966-6362(00)00048-5
Beaudoin L Zabjek KF Leroux MA Coillard C Rivard CH Acute systematic and variable postural adaptations induced by an orthopaedic shoe lift in control subjects Eur Spine J 1999 8 40 45 10190853 10.1007/s005860050125
Travell JG Simons DG Chapter 4, Quadratus Lumborum Muscle Myofascial Pain and Dysfunction The Trigger Point Manual The Lower Extremities 1999 2 2 Williams & Wilkens: Baltimore 104
Giles LGF Lumbosacral facetal 'joint angles' associated with leg length inequality Rheumatology and Rehabilitation 1981 20 233 238 6458081
Giles LGF Taylor JR Lumbar spine structural changes associated with leg length inequality Spine 1982 7 159 162 7089692
Levangie PK The association between static pelvic asymmetry and low back pain Spine 1999 24 1234 42 10382251 10.1097/00007632-199906150-00011
Knutson G Incidence of foot rotation, pelvic crest unleveling, and supine leg length alignment asymmetry, and their relationship to self-reported back pain J Manipulative Physiol Ther 2002 24 e1 10.1067/mmt.2002.121414
Fann AV The prevalence of postural asymmetry in people with and without chronic low back pain Arch Phys Med Rehabil 2002 83 1736 8 12474179 10.1053/apmr.2002.35653
Andersson GB Epidemiological features of chronic low-back pain Lancet 1999 354 581 85 10470716 10.1016/S0140-6736(99)01312-4
Bronfort G Goldsmith CH Nelson CF Boline PD Anderson AV Trunk exercise combined with spinal manipulative or NSAID therapy for chronic low back pain: A randomized, observer-blinded clinical trial J Manipulative Physiol Ther 1996 19 570 82 8976475
Friberg O Letter-to-the-editor Spine 1992 17 458 460 1533733
Reid DC Smith B Leg length inequalty: A review of etiology and management Physiotherapy Canada 1984 36 177 182
Kujala UM Friberg O Aalto T Kvist T Osterman K Lower limb asymmetry and patellofemoral joint incongruence in the etiology of knee exertion injuries in athletes Int J Sports Med 1987 8 214 20 3623784
Kujala UM Kvist M Osterman K Friberg O Aalto T Factors predisposing army conscripts to knee exertion injuries incurred in a physical training program Clin Orthop Rel Res 1986 210 203 12
Subotnick SI Limb length discrepancies of the lower extremity (The short leg syndrome) JOSPT 1981 3 11 16
Gibson PH Papaioannou T Kenwright J The influence on the spine of leg-length discrepancy after femoral fracture J Bone Joint Surg (Br) 1983 65 584 7 6643562
Edeen J Sharkey PF Alexander AH Clinical significance of leg-length inequality after total hip arthroplasty Am J Orthop 1995 24 347 351 7788314
White TO Dougall TW Arthroplasty of the hip. Leg length is not important J Bone Joint Surg (Br) 2002 84-B 335 8 12002488 10.1302/0301-620X.84B3.12460
Paravizi J Sharkey PF Bissett GA Rothman RH Hozack WJ Surgical treatment of limb-length discrepancy following total hip arthroplasty J Bone Joint Surg 2003 85-A 2310 17 14668499
Yrjönen T Hoikka V Poussa M Österman K Leg-length inequality and low-back pain after Perthes' disease: A 28–47 – year follow-up of 96 patients J Spinal Disord 1992 5 443 447 1490042
Papaioannou T Stokes I Kenwright J Scoliosis associated with limb-length inequality J Bone Joint Surg 1982 64-A 59 62 7054204
Friberg O Leg length asymmetry in stress fractures J Sports Med 1982 22 485 488
Korpelainen R Orava S Karpakka J Siira P Hulkko A Risk factors for recurrent stress fractures in athletes Am J Sports Med 2001 29 304 10 11394600
Yen ST Andrew PD Cummings GS Short-term effect of correcting leg length discrepancy on performance of a forceful body extension task in young adults Hiroshima J Med Sci 1998 47 139 43 9973739
Murrell P Cornwall MW Doucet SK Leg-length discrepancy: effect on the amplitude of postural sway Arch Phys Med Rehabil 1991 72 646 8 1859258
Goel A Loudon J Nazare A Rondinelli R Hassanein K Joint moments in minor limb length discrepancy: A pilot study Am J Orthop 1997 26 852 6 9413588
Brand RA Yack JH Effects of leg length discrepancies on the forces at the hip joint Clin Ortho Rel Res 1996 333 172 180 10.1097/00003086-199612000-00017
Perttunen JR Anttila E Socergard J Merikanto J Komi PV Gait asymmetry in patients with limb length discrepancy Scand J Med Sci Sports 2004 14 49 56 14723788 10.1111/j.1600-0838.2003.00307.x
Kaufman KR Miller LS Sutherland DH Gait asymmetry in patients with limb-length inequality J Ped Orthop 1996 16 144 150 10.1097/00004694-199603000-00002
White SC Gilchrist LA Wilk BE Asymmetric limb loading with true or simulated leg-length differences Clin Orthop 2004 421 287 292 15123962
Gurney B Mermier C Robergs R Gibson A Rivero D Effects of Limb-Length Discrepancy on Gait Economy and Lower-Extremity Muscle Activity in Older Adults J Bone Joint Surg Am 2001 83 907 915 11407800
Triano JJ Objective electromyographic evidence for the use and effects of lift therapy J Manipulative Physiol Ther 1983 6 13 16 6222127
Gofton JP Persistent low back pain and leg length disparity J Rheumatol 1985 12 747 750 2932553
Moseley CF Leg length discrepancy and angular deformity of the lower limbs Lovell and Winter's Pediatric Orthopaedics 1996 4 Philadelphia: Lippencott-Raven 877
Kakushima M Miyamoto K Shimizu K The effect of leg length discrepancy on spinal motion during gait Spine 2003 28 2472 6 14595166 10.1097/01.BRS.0000090829.82231.4A
|
16026625
|
PMC1232860
|
CC BY
|
2021-01-04 16:38:23
|
no
|
Chiropr Osteopat. 2005 Jul 20; 13:11
|
utf-8
|
Chiropr Osteopat
| 2,005 |
10.1186/1746-1340-13-11
|
oa_comm
|
==== Front
CytojournalCytoJournal1742-6413BioMed Central London 1742-6413-2-131615014110.1186/1742-6413-2-13CommentaryHodgkin lymphoma: flow me? Beaty Michael W [email protected] Kim R [email protected] Wake Forest University School of Medicine Department of Pathology Medical Center Boulevard Winston-Salem, North Carolina 27157-1072 USA2005 8 9 2005 2 13 13 5 8 2005 8 9 2005 Copyright © 2005 Beaty and Geisinger; licensee BioMed Central Ltd.2005Beaty and Geisinger; 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.
Combining fine needle aspirate cytology with flow cytometry immunophenotyping for the rapid diagnosis of lymphoproliferative lesions is commonplace practice in many institutions. Yet, a definitive diagnosis of Hodgkin lymphoma in many cases remains elusive, requiring subsequent tissue biopsy confirmation. In this issue of CytoJournal, Hernandez et al explore the potential role of using the increased CD4/CD8 T-cell ratio in lymph node fine needle aspiration specimens as a specific feature in diagnosing Hodgkin lymphoma. CD4/CD8 T-cell ratio comparisons are made with cytomorphologic diagnoses of reactive, atypical, non-Hodgkin lymphoma, and Hodgkin lymphoma cases.
==== Body
Fine needle aspiration (FNA) cytomorphology in conjunction with flow cytometric (FC) immunophenotyping has become a reliable and accurate method for the diagnosis and classification of many lymphoproliferative disorders. The combined modality approach often allows for confirmation of diagnosis and subtyping of lymphoma, and in many cases obviates the necessity for a more invasive open biopsy for many patients with lymphadenopathy [1]. This is especially true in suspected cases of non-Hodgkin lymphomas (NHL). However, even with the collaborative expertise that we and others routinely employ to reach a consensus diagnosis between the cytopathologist interpreting the FNA material and the hematopathologist signing-out the FC data, the diagnosis of Hodgkin lymphoma (HL) through this method remains difficult. A FNA diagnosis of HL based purely on cytomorphology is often followed by tissue biopsy [2].
One of the major difficulties encountered with combined FNA/FC diagnosis of HL is identifying sufficient numbers of the neoplastic cells for subsequent phenotypic analysis. In most instances, too few malignant cells for FC can be obtained from the tumors, either by FNA or biopsy material. This is due, in large part, to the cellular makeup of HL, in which the inflammatory background cells outnumber the malignant RS cells and variants by 10–100 fold. In addition, the neoplastic cells exhibit high cellular fragility, and are associated with a fibrous stroma interfering with cellular dispersion. This has led to Herculean gating efforts to isolate the RS cells and phenotype them with CD30 and CD15 [3,4]. Many other studies have looked at the components of the normal background T-lymphocyte populations present in HL, where it has long been known that there is a relative preponderance of CD4-positive T-cells, with an increased helper/cytotoxic CD4/CD8 ratio by immunohistochemistry [5-7]. We know that the RS cells and variants within the lymph node affect the surrounding cells, which results in alterations of the proportion between T and B lymphocytes and the activation of these cells, and that the relative proportion of the background lymphocytes differs depending upon the histologic subtype of the HL [8]. While the nature and composition of the background reactive lymphocytes is helping to increase our basic understanding of the pathogenesis of the disease, it can hardly be used as a specific diagnostic feature.
In the current issue of Cytojournal, Hernandez et al compare 85 cases of combined FNA/FC lymph nodes with an increased CD4/CD8 ratio (>4), and demonstrate no definitive quantitative differences in T-lymphocyte distribution between HL, NHL, and benign reactive lymph nodes. They report that the average CD4/CD8 ratio was essentially identical in reactive, HL, and NHL cases. This should not be surprising given the extraordinary variation in lymphocyte components in many lymphoproliferative processes. T-cell rich large B-cell lymphoma and nodular lymphocyte predominant HL are just two tumors which quickly come to mind. Both of these tumors can morphologically mimic classical HL particularly on FNA material, with only a minor population of malignant RS-like cells within a sea of reactive small lymphocytes. Phenotypic studies confirm that CD4-positive cells may exceed the CD8-positive background lymphocytes [10,11].
In a recent report by Ravoet, et al, FC was combined with lymph node biopsy to help discriminate HL, NHL, carcinoma, and reactive diagnoses [4]. Median CD4/CD8 ratios in HL were 5.0 (range 1.6–10.4), which was similar to that seen in granulomatous lymph nodes (median 3.2, range 1.5–11.8), and reactive hyperplasia (median 2.71, range 0.4–18.6). They concluded that given the large scattering of the CD4/CD8 ratios within each group, the use of the CD4/CD8 ratio is not recommended in clinical practice in order to diagnosis HL.
A quick review over the past year of lymph node FNAs performed at our institution reveals a similar wide scattering of ratios. We identified 6 cases of FNA suspected HL with a median CD4/CD8 ratio of 9.7 (range 0.3–42.5). Subsequent nodal biopsies confirmed HL in all cases. The CD4/CD8 ratio was less than 4 in half of these cases. Interestingly, one patient with an HIV-associated HL predictably had a reverse ratio, another potential pitfall. Equally important though were the six additional cases of suspected HL by FNA in which the specimens received for FC were inadequate, containing too few cells to phenotype. Thus, we rely exclusively on the morphologic features present within the FNA smears, combined with confirmatory immunohistochemical studies performed on cell block material to diagnosis HL with confidence. Otherwise, a nodal biopsy is and should be recommended [1].
Integration of clinical and biological features is one of the basic tenets of the WHO classification of lymphoproliferative disorders. While combined FNA/FC is extremely useful in the diagnosis and subclassification of some lymphoproliferative disorders, it is not without limitation and is not the end of the diagnostic line. Until we know more about the characteristics of the T-cell milieu of HL and can point to a more specific phenotype which can be detected by FC, we must continue to rely on other ancillary tools for a definitive HL diagnosis. The work of Hernandez et al is important to remind us that an increased CD4/CD8 ratio is only one phenotypic clue present in HL, and a non-specific clue at that.
Note
Corresponding article: Hernandez O, Oweity T, Ibrahim S: Is an Increase in CD4/CD8 T-cell ratio in lymph node fine needle aspiration helpful for diagnosing Hodgkin lymphoma? Cytojournal 2005, 2:14 [9].
==== Refs
Meda BA Buss DH Woodruff RD Cappellari JO Rainer RO Powell BL Geisinger KR Diagnosis and subclassification of primary and recurrent lymphoma. The usefulness and limitations of combined fine-needle aspiration cytomorphology and flow cytometry Am J Clin Pathol 2000 113 688 99 10800402 10.1309/0Q7F-QTGM-6DPD-TLGY
Moreland WS Geisinger KR Utility and outcomes of fine-needle aspiration biopsy in Hodgkin's disease Diagn Cytopathol 2002 26 846 8 10.1002/dc.10104
Dunphy CH Contribution of flow cytometric immunophenotyping to the evaluation of tissues with suspected lymphoma Cytometry 2000 42 296 306 11025488 10.1002/1097-0320(20001015)42:5<296::AID-CYTO7>3.0.CO;2-X
Ravoet C Demartin S Gerard R Dehon M Peny MO Petit B Delannoy A Husson B Contribution of flow cytometry to the diagnosis of malignant and non-malignant conditions in lymph node biopsies Leukemia & Lymphoma 2004 45 1587 93 15370210 10.1080/10428190310001609906
Forni M Hofman FM Parker JW Lukes RJ Talor CR B- and T-lymphocytes in Hodgkin's disease. An immunohistochemical study utilizing heterologous and monoclonal antibodies Cancer 1985 55 728 37 3155642
Knowles DM T-lymphocyte subpopulations in B-cell derived non-Hodgkin's lymphomas and Hodgkin's disease Cancer 1984 54 644 651 6234982
Gattringer G Griel R Radaszkiewicz T Hubert H In situ quantification of T-cell subsets, NK-like cells and macrophages in Hodgkin's disease: quantity and quality of infiltration density depnds on histopathological subtypes Blut 1986 53 49 58 3487362 10.1007/BF00320582
Pituch-Noworolska A Drabik G Kacinska E Klekawka T Lymphocyte populations in lymph nodes in different histological types of Hodgkin's disease in children Acta Haematol 2004 112 129 35 15345894 10.1159/000079723
Hernandez O Oweity T Ibrahim S Is an increase in CD4/CD8 T-cell ratio in lymph node fine needle aspiration helpful for diagnosing Hodgkin lymphoma? Cytojournal 2005 2 14 16153296 10.1186/1742-6413-2-14
Axdorph U Powit-MacDonald A Sjøberg J Grimfors G Bjørkholm M T-cell rich B-cell lymphoma – diagnostic and therapeutic aspects APMIS 2002 110 379 90 12076255 10.1034/j.1600-0463.2002.100503.x
Chan WC Cellular origin of nodular lymphocyte-predominant Hodgkin's lymphoma: immunophenotypic and molecular studies Semin Hematol 1999 36 242 52 10462324
|
16150141
|
PMC1232861
|
CC BY
|
2021-01-04 16:36:22
|
no
|
Cytojournal. 2005 Sep 8; 2:13
|
utf-8
|
Cytojournal
| 2,005 |
10.1186/1742-6413-2-13
|
oa_comm
|
==== Front
Int Semin Surg OncolInternational seminars in surgical oncology : ISSO1477-7800BioMed Central London 1477-7800-2-171613893110.1186/1477-7800-2-17ResearchIncidence and risk factors of venous thromboembolism (VTD) in patients with amyloidosis Srkalovic Gordan [email protected] Marte G [email protected] Steven R [email protected] Kandice [email protected] Mohamad A [email protected] Sparrow Regional Cancer Center, Lansing, Michigan, USA2 Helgelandssykehuset, Avdeling Sandnessjøen, Sandnessjøen, Norway3 Nuvelo Inc, Sunnyvale, California, USA4 Department of Pathology, Cleveland Clinic Foundation, Cleveland, Ohio, USA5 Cleveland Clinic Myeloma Research Program, Cleveland Clinic Foundation, Cleveland, Ohio, USA2005 2 9 2005 2 17 17 1 8 2005 2 9 2005 Copyright © 2005 Srkalovic et al; licensee BioMed Central Ltd.2005Srkalovic 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
Coagulation problems in amyloidosis are historically associated with bleeding tendencies (mostly Factor X abnormalities). Increased clotting was observed in isolated cases diagnosed with low-grade disseminated intravascular coagulation (DIC). Problem of venous thromboembolic disaease (VTD) in amyloidosis was not systematically investigated.
Methods
We evaluated frequency of VTD and risk factors for VTD in 56 consecutive amyloidosis patients with a documented disease evaluated and followed up at our Center from 1991–2001. Data was collected in 5 categories: (a) demographics, (b) disease and treatment, (c) thrombosis case information, (d) major risk factors for thrombosis and (e) baseline laboratory data. Univariable correlates of VTD were assessed using Kaplan-Meier analysis and Cox proportional hazards analysis.
Results
Mean age of the patients was 67 (years range 21 – 83). Male/female percentage ratio was 70/30. 29 % of the patients had high creatinine level (> 1.4 mg/dl). Personal or family history of VTD was recorded in 2 and 0 % of patients, respectively. Known hypercoagulable state was present in 1 patient (2%). 8 % of patients were smokers. Of 56 patients, 6 developed VTD (11%). Median time from diagnosis of amyloidosis to VTD was 12.5 month (range 1–107). Treatment was given within a median of 1 month (range 0–4) from the development of thrombosis. Only sites of VTD were lower extremities. No cases were associated with I.V. line. 1 case (17 %) was identified postoperatively. We identified several univariable correlates of VTD in amyloid patients, including greater age at diagnosis (HR-2.99, P = .041), personal history of DVT (HR-47.7, P = .006) and immobility (HR-11.78, P = .006). Presence of circulating serum M-protein had protective role in our analysis (HR-.08, P = .031). There was no correlation with the type of treatment patients were receiving.
Conclusion
Risk for thromboembolic diseases in patients with amyloidosis is similar to one previously described for multiple myeloma. Additional studies with higher number of thromboembolic events could help to further elucidate risk factors for VTD in this population of patients.
==== Body
Background
amyloidosis is characterized by organ deposition of fibrillar substances of different types that have a homogeneous eosinophilic appearance on light microscopy [1,2]. The diagnosis is confirmed by demonstration of an apple-green birefringence under polarized light of specimens stained with Congo red. Primary amyloidosis (AL) should be considered in any patient who has a monoclonal protein in the serum or urine and refractory congestive heart failure, nephrotic syndrome, sensorimotor peripheral neuropathy, carpal tunnel syndrome, orthostatic hypotension, or steatorrhea. At diagnosis, approximately one-third have nephrotic syndrome, one-fifth have carpal tunnel syndrome, one-fourth have congestive heart failure, and about 15% have peripheral neuropathy [1,2].
Bleeding problems due to vascular deposition of amyloid resulting in vascular fragility, deficiencies of clotting factors (particularly Factor X) or enhanced fibrinolytic activity often complicate this disease [3]. Acquired deficiency of Factor X is the most common coagulation factor deficiency identified in patients with amyloidosis [4]. Thrombosis, although less common, also can occur. Some case reports described thromboembolic complications of amyloidosis [5,6]. Association of nephrotic syndrome, well-documented complication of systemic amyloidosis, and renal vein thrombosis are well documented [7,8]. Cardiac amyloidosis produces cases of atrial electromechanical dissociation with resulting atrial arrhythmias and formation of atrial clots [9]. Various pathogenic factors have been proposed to explain abnormal hemostasis in patients with amyloidosis. Recently, impairment of the thrombin-antithrombin pathway, in association with the low antithrombin biological activity was recognized to have possible pathogenic role in hypercoagulability of amyloidosis patients [10].
However, the problem of venous thromboembolic disease (VTD) in this disease was not systematically investigated. We evaluated the frequency of VTD and risk factors for VTD in 56 consecutive amyloidosis patients with a documented disease evaluated and followed up at our Center from 1991–2001.
Methods
We reviewed complete medical records of all patients with amyloidosis who were prospectively entered in the database. Fifty-six patients were identified. They were evaluated and followed up by staff oncologists in the Department of Hematology and Oncology, Myeloma Research Program at the Cleveland Clinic Foundation in the period 1991–2001.
The Cleveland Clinic Foundation (CCF) is an urban tertiary referral center and a teaching hospital. Our proposal was submitted to the CCF Institutional Review Board, and we were granted approval, based on an expedited review. The requirement for obtaining informed consent was waived based on the study designs (standard medical record based review) and patient anonymity. The study did not require any third party funding.
The patients' computerized inpatient and outpatient medical records were initially reviewed by one of the co-investigators (CMG). Additional review was done independently by a second co-investigator (SG). If electronic data were incomplete, the hard copies of charts were also reviewed. When necessary, missing information was sought from referring providers. However, in a number of cases, this was unavailable.
All cases of VTD of extremities were diagnosed by Duplex ultrasound. Term immobility was used for patients who were hospitalized or bed bound for more than 7 days at the time of VTD diagnosis.
Personal history of VTD was defined as a history of documented thromboembolism prior to the diagnosis of PCD. Patients were screened for a hypercoagulable state only when they developed what appeared to be a vascular event that needed further investigation.
Data was collected in 5 categories: (a) demographics, (b) disease and treatment, (c) thrombosis case information, (d) major risk factors for thrombosis and (e) baseline laboratory data. Data were manually recorded onto researcher-developed data collection forms and subsequently entered into a computerized database.
Statistical analysis
Categorical variables are summarized as frequencies and percentages and continuous variables as the mean, standard deviation, median, and range. Time until development of a VTD was estimated using the Kaplan-Meier method. Follow-up was calculated from the date of diagnosis to the date of either the first VTD or the date of final follow-up visit. Cox proportional hazards analysis was used to identify univariable correlates of thromboembolic event. If Cox analysis could not be done due to lack of variability, then the log-rank test was performed instead.
There were 8 possible risk factors for VTD: personal and family history of VTD, known hypercoagulable state, non-plasma cells related concomitant neoplasm's, immobility, estrogen treatment, smoking, and use of indwelling catheters at any time from initial diagnosis.
All statistical analyses were two-sided; P < 0.05 was used to indicate statistical significance. Analyses were performed using SAS® software, version 6.12.
Results
Patients' characteristics
Data was available on 56 patients (Table 1.). 50 (89%) had primary amyloidosis (AL), 1 (2%) had kappa light chain disease, and in 5 (9%) patients type could not be identified. One patient had both AL and familial amyloidosis. 37 (66%) had confirmed cardiac involvement by biopsy or alternatively, typical findings on Echocardiogram. Confirmed (biopsy) renal involvement was found in 18 (32%) of patients. Localized amyloidosis was identified in 6 (11%) of patients. Two cases were pulmonary, 2 gastrointestinal, and one case was localized to skin and genitourinary tract, respectively. 39 (70%) were men and 17 (30%) were women. Average age was 67 years (range 21 – 83). Of the heavy chain types found in patients, 75% were IgG, 17% IgA, 1% IgM. 44/56 (78.5%) of patients did not have heavy chain M-protein in the serum.
Table 1 Clinical characteristics of amyloidosis patients at presentation (N = 56)
Characteristics
Male/Female 39/17
Age (years)
Median 67
Range (21–83)
Serum M-Protein (g/L)1 (N = 18)
Mean 0.5
Median 0.2
Range (0–4.0)
24-hour Urine Protein (g)1 (N = 47)
Mean 3.2
Median 0.4
Range (0–26.9)
BMPC* (%)2 (N = 53)
Mean 8
Median 5
Range (0–50)
B2 Microglobulin (mg/dl)1 (N = 47)
Mean 3.4
Median 2.7
Range (0.3–18.4)
Serum albumin (g/L)1 (N = 53)
Mean 3.4
Median 3.6
Range (1.0–5.4)
Immunoglobulin type Heavy Chain (N = 12)
IgA (%) 2 (17)
IgG (%) 9 (75)
IgM (%) 1 (8)
Unknown 0
Light Chain (N = 25)
λ 18 (72)
κ 7 (28)
History of VTD Personal
Yes (%) 1 (2)
No (%) 53 (94)
Unknown (%) 2 (4)
Family
No (%) 52 (93)
Unknown (%) 4 (7)
1 CCF Laboratory reference range, normal values:
Serum M-protein: 0 g/dl
Urine Protein: 0–0.15 g/24°
β2 Macroglobulin: 0.3–1.9 mg/dl
Serum Albumin: 3.5–5.0 g/L
2 Bone Marrow Plasma Cells
The light chain was lambda (λ) in 18 (32% of all patients) and kappa (κ) in 7 patients (13%). The remainder did not have identifiable light chain (31, 55%). All patients had amyloid deposits by Congo Red stain. The bone marrow biopsy results were available in 53 patients (95%) at the time of diagnosis. The median percentage of bone marrow occupied by plasma cells was 5% (range 0.0 – 50%). Serum creatinine level was elevated in 29% of all patients. Albumin was low (< 3.5 gm/dl) in 45% patients overall. Beta-2-microglobulin was high (>1.9) in the majority of patients (70%). Bone survey was negative for lytic lesions in all cases.
Out of 56 patients with amyloidosis, 1 (2%) patients had personal and no one had family history of thromboembolism. 1/56 were diagnosed with hypercoagulable state (antithrombin deficiency). Patients were treated with multiple regimens that included: steroids (36%), α-interferon (23%), vincristine (16%), liposomal doxorubicin (16%), melphalan (9%), thalidomide (2%) and rituximab (2%).
The five different drug combinations used in the treatment of amyloid patients are presented in Table 2. No patients were treated with high-dose chemotherapy with peripheral stem cell support (HDC-PSC).
Table 2 Combination chemotherapies used in the treatment of amyloidosis
Type of Treatment N %
HD steroids1 20 36
DVd2 9 16
MP3 5 8
CP4 1 2
TD5 1 2
N – Number of patients
% – Percentage of patients treated with particular regimen
1 High-dose Dexamethasone
2 Doxil®, Vincristine, Dexamethasone
3 Melphalan, Prednisone
4 Cytoxan, Prednisone
5 Thalidomide, Dexamethasone
Patients with VTD
Cases of thromboembolism were diagnosed by Duplex ultrasound. There was no evidence of pulmonary embolism or atrial clots in this group of patients. Of 56 patients reported in this paper 6 (11%) had one or more episodes of symptomatic, objectively documented VTD after the diagnosis of amyloidosis (Table 3). All cases were of lower extremities. The time from diagnosis to development of VTD was 12.5 months (range 1–107). Median time from last treatment to VTD event was 1 month (0–4). Out of 6 patients with VTD 17% (1 case) had a personal history of thromboembolism. No patients with VTD had a history of hypercoagulable state, smoking, or oral estrogen use. Four (67%) patients with VTD had prolonged immobility and 1 (17%) had indwelling catheter at some point during the disease. One patient (17%) developed VTD within 6 weeks from surgery.
Table 3 Clinical characteristics of amyloidosis (N = 6) patients with VTD
Characteristics
Male/Female 4/2
History of VTD
Personal
Yes 1
No 5
Family
No 6
Serum M-Protein1
None 3
Yes 2
Unknown 1
24-hour Urine Protein1
<0.15 gr 3
>0.15 gr 2
Unknown 1
BMPC2
≤ 5% 3
> 5% 1
Unknown 2
B2 Microglobulin1
< 1.9 mg/L 1
> 1.9 mg/L 4
Unknown 1
Serum albumin1
< 3.5 g/L 3
3.5 – 5.0 g/L 3
Immunoglobulin type
None 4
IgG 1
IgM 1
None 4
κ 2
WBC count
4 – 11 × 106/L 6
Platelet count
150 – 400 × 106/L 6
1CCF Laboratory reference range, normal values:
Serum M-protein: 0 g/dl
Urine Protein: 0–0.15 g/24°
β2 Macroglobulin: 0.3–1.9 g/dl
Serum Albumin: 3.5–5.0 g/L
2 Bone Marrow Plasma Cells
Univariable correlates of VTD in amyloidosis patients are presented in Table 4. A large number of factors were entered into the model including the following: gender, age at diagnosis, immunoglobulin type, personal history of VTD, hypercoagulable state, non-plasma cell related concomitant neoplasms, smoking, immobility, indwelling catheters, estrogen treatment, percentage of plasma cells in the bone marrow biopsy, creatinine level, corrected serum calcium level, serum albumin, white blood cells and platelet count, hemoglobin level, β 2-microglobulin, serum M protein level, 24 hour urine protein level, as well as individual agents and combination of chemotherapeutic agents.
Table 4 Identifying univariable correlates of VTD in amyloidosis (N-56)
Variable (Reference)
HR2 95% CI2 P1
Age at Diagnosis (per 10 year increase) 2.99 1.05–8.56 0.0411
Personal History of VTD (Yes/No) 47.7 2.98–766 0.0061
Immobile (Yes/No) 11.78 2.02–68.8 0.0061
Serum M-protein
Per 1 mg/dl increase 0.43 0.05–3.70 0.45
Normal/Abnormal 12.5 1.25–100 0.0311
1 significant (P < 0.05)
2 HR – Hazard Ratio; CI – Confidence Interval
Among these, personal history of VTD (HR-47.7; P = .006, CI-2.98–766), immobility (HR-11.78; P = .006, CI-2.02–68.8), absence of measurable M-protein (HR-12.5; P = .031, CI-1.25–100), and age at diagnosis (HR-2.99, P = .041, CI-1.05–8.56) were identified as univariable correlates of VTD in amyloidosis patients (Table 4.) Multivariable analysis could not be done due to low number of events.
Discussion
This study shows increased incidence of VTD (11% at 10 years review) in amyloidosis patients as compared to the general population (0.1% per person/per year) [11]. This is similar to frequency of VTD found in multiple myeloma (MM) (10%) [12], and slightly higher than in patients with monoclonal gammopathy with undetermined significance (MGUS) (7.5%) [12]. Previously, only case reports of thromboembolism in patients with amyloidosis were presented in the literature [5,6,9,13].
Studies presented herein have significant limitations. It is highly dependent on the quality of historical data. Absence of a potential risk factor or patient characteristics in the medical records is assumed to mean that the factor is absent. The statistical analysis is limited by the fact that several of the variables that were considered as correlates of VTD were not baseline characteristics. Ideally, these variables would be analyzed as time-dependent covariates. However, timing of these variables was not known. It is only known that they occurred before development of the event in patients with VTD or before the last follow-up among patients who did not develop VTD yet.
Results of the study presented herein point to strong correlation between some of the established risk factors for VTD (personal history and immobility) and thromboembolic events in patients with amyloidosis. Other well-known VTD risk factors (estrogen use, concomitant solid malignancies, and use of indwelling catheters) did not seem to be of importance in our patients.
Data about family history of VTD is missing in more than 10% of patients. Additionally, existing data seem to be inadequate. From available records it seems that none of the 56 patients had a family history of VTD. Although this could be due to the low number of patients a more probable explanation is poor history taking. Therefore; family history was not included in univariable analysis. Types of treatment (individual drugs or chemotherapy combinations) do not correlate with VTD. It is important to mention that none of the patients included in this study had treatment with high dose chemotherapy and stem cell support.
One of the univariable correlates of VTD identified in our group of amyloidosis patients is increased age. Risk of VTD in these patients is increasing 3 times with every 10-year increase in age. A number of studies support an association between increasing age and a higher incidence of venous thromboembolism [14-16]. Patients over age 40 are at a significantly increased risk compared with younger patients [17]. Risk continues to increase with increasing age, approximately doubling with each decade after age of 40 [18].
It seems that in our amyloidosis patients, VTD age risk is in excess of that seen in general population. It does not seem that this increased risk correlates with more aggressive disease, since age was not found to be an important prognostic factor for survival in primary systemic amyloidosis in multiple studies [19-21]. However, other types, such as senile with normal transthyretin (ATTR), amyloidosis associated with Alzheimer's disease (Aβ), and islet amyloid polypeptide deposits (AIAPP) are very strongly associated with both old age and the aging process.
Another significant risk factor of VTD in our patients was absence of circulating monoclonal protein. This is a characteristic for types of disease other than primary amyloidosis (AL) including secondary (AA), familial with genetically variant transthyretin (ATTR), amyloidosis associated with dialysis (Aβ2M), as well as Aβ and AIAPP forms. Again, these results point to possible increase incidence of VTD in the forms other than primary.
Human serum amyloid A (SAA) is a precursor protein in inflammation-associated secondary amyloidosis (AA). It could interfere with the coagulation process through interaction with heparin/heparan sulfate at the specific binding site (C-terminal residue 78 to 104) [22]. SAA also seems to play a role in atherogenesis, as well as in thrombus formation at the vascular injury site [23,24]. Secondary amyloidosis (AA) is manifested in patients with underlying chronic inflammation (rheumatoid arthritis, ankylosing spondylitis, inflammatory bowel disease, tuberculosis, osteomyelitis, Familial Mediterranean Fever) [25-27]. It is driven by different cytokines, including tumor necrosis factor (TNF-α), interleukin-6 (IL-6), and interleukin-1 (IL-1) [28]. IL-6 was found to promote coagulation without affecting fibrinolysis [29]. It activates coagulation cascade through tissue factor stimulation and increased transcription of factor VIII, up regulates transcription of fibrinogen, increases von Willebrand's factor, and decreases protein S [29]. Is there any correlation between SAA or any other circulating non-monoclonal protein precursors of amyloid and VTD will have to be elucidated in future studies. It is still uncertain if proposed increased incidence of thromboembolism is the result of the very process of deposition of the abnormally folded protein outside cells, consequential organ damage, or alternatively, underlying processes leading to amyloid deposits. However, it does not seem to be related to the type of treatment used in this disease. This is in accordance with our own findings in the patients with multiple myeloma and MGUS [12].
Conclusion
Our analysis confirmed that well-known risk factors for thromboembolism including personal history of VTD and immobilization could represent confounding factors in amyloidosis patients, too. However, additional correlates have to be considered, including type of disease and patients' age.
In our opinion, these patients need to be closely monitored for development of VTD. Standard thromboembolic prophylaxis should be used with extreme caution, considering bleeding tendency in the patients with amyloidosis.
In our opinion development of prospective studies dealing with venous thromboembolism in amyloidosis patients is of utmost importance. We will need to define pathophysiological mechanisms responsible in order to employ the most effective prophylactic measures.
Competing interests
The author(s) declare they have no competing interest.
Authors' contributions
GS have made significant contribution to conception and design of the study, analysis and interpretation of the data and drafting the manuscript
MGC have been involved in collection, analysis, interpretation of the data and critical revision of the manuscript
SRD have been involved in drafting and revising the manuscript
KKM have been involved in revising manuscript for important intellectual content
MAH have made significant contribution to concept and design of the study, revision of the manuscript and have given final approval of the version to be published.
All authors read and approved manuscript
Acknowledgements
Authors would like to thank Miss Marion McElhaney for valuable technical assistance.
==== Refs
Gertz MA Kyle RA Primary systemic amyloidosis – a diagnostic primer Mayo Clinic 1989 Proc 64 1505
Kyle RA Gertz MA Primary systemic amyloidosis: clinical and laboratory features in 474 cases Sem Hematol 1995 32 45
Glaspy JA Hemostatic abnormalities in multiple myeloma and related disorders Hem/Onc Clin North Amer 1992 6 1301 1315
Greipp PR Kyle RA Bowie EJ Factor-X deficiency in amyloidosis: a critical review Am J Hematol 1981 11 443 450 7036713
Catovsky D Ikoku NB Pitney WR Galton DAG Thromboembolic Complications in Myelomatosis Brit Med J 1970 3 438 439 5454323
Tinaztepe K Buyan N Tinaztepe B Akkok N The association of nephrotic syndrome and renal vein thrombosis: a clinicopathological analysis of 8 pediatric patients Turk J Ped 1989 31 1 18
Kim V Spandorfer J Epidemiology of venous thromboembolic disease Emerg Med Clin N Amer 2001 19 839 59 10.1016/S0733-8627(05)70221-2
Cameron JS Coagulation and thromboembolic complications in the nephrotic syndrome Adv Nephr Necker Hosp 1984 13 75 114
Dubrey S Pollak A Skinner M Falk RH Atrial thrombi occuring during sinus rhythm in cardiac amyloidosis: evidence for atrial electromechanical dissociation Br Heart J 1995 74 541 544 8562243
Gamba G Montani N Anesi E Palladini G Lorenzutti F Perfetti V Merlini G Abnormalities in thrombin-antithrombin pathway in AL amyloidosis amyloid 1999 6 273 277 10611948
Hirsch J Lee A How we diagnose and treat deep vein thrombosis Blood 2002 99 3102 3110 11964271 10.1182/blood.V99.9.3102
Srkalovic G Camerom MG Rybicki L Deitcher SR Kottke-Marchant K Hussein MA Monoclonal gammopathy of undetermined significance and multiple myeloma are associated with an increased incidence of venothromboembolic disease Cancer 2004 101 558 15274069 10.1002/cncr.20405
Baier J Neumann H Grossmann R Ricken D Primary amyloidosis and thrombophlebitis as complications of Bence-Jones plasmacytoma (lambda type) Medizinishe Klinik 1991 86 53 57
Gillum RF Pulmonary embolism and thrombophlebitis in the United States, 1970–1985 Am Heart J 1987 114 1262 1264 3673898 10.1016/0002-8703(87)90212-2
Gjores JE The incidence of venous thrombosis and its sequelae in certain districts in Sweden Acta Chirurgica Scandinavica 1987 111 16 24
Nicolaides AN Irving D Nicolaides AN Clinical factors and the risk of deep venous thrombosis Thromboembolism Etiology, Advances in Prevention and Management 1975 Baltimore: University Park press 193 204
Anderson FA JrWheeler HB Goldberg RJ A population-based perspective of the hospital incidence and case-fatality rates of deep venous thrombosis and pulmonary embolism: The Worcester DVT Study Arch Int Med 1991 151 933 938 2025141 10.1001/archinte.151.5.933
Anderson FA JrWheeler HB Venous thromboembolis. Risk Factors and Prophylaxis Clin Chest Medic 1995 16 235 251
Kyle RA Greipp PR O'Fallon WM Primary systemic amylodosis: multivariate analysis for prognostic factors in 168 cases Blood 1986 1 220 224 3719098
Pardanani A Witzig TE Schroeder G McElroy EA Fonseca R Dispenzieri A Lacy MQ Lust JA Kyle RA Greipp PR Gertz MA Rajkumar SV Circulating peripheral blood plasma cells as a prognostic indicator in patients with primary systemic amyloidosis Blood 2003 101 827 830 12393530 10.1182/blood-2002-06-1698
Dipsenzieri A Lacy MQ Kyle RA Therneau TM Larson DR Rajkumar SV Fonseca R Greipp PR Witzig TE Lust JA Gertz MA Eligibility for hematopoetic stem-cell transplantation for primary systemic amyloidosis is a favorable prognostic factor for survival J Clin Oncol 2001 19 3350 3356 11454882
Ancain JB Kisilevsky R The heparin/heparan sulfate-binding site on aposerum amyloid A: implicationfor the therapeutic intervention in amyloidosis Int J Exp Clin Invest 1999 6 37 47
Meek RL Urieli-Shoval S Benditt EP Expression of apolipoprotein serum amyloid A mRNA in human atherosclerotic lesions and cultured vascular cells: implication for serum amyloid A function Proc Natl Acad Sci USA 1994 91 3186 3190 8159722
Urieli-Shoval S Linke RP Matzner Y Expression and function of serum amyloid A, a major acute-phase protein, in normal and disease state Curr Opin Hematol 2000 7 64 69 10608507 10.1097/00062752-200001000-00012
Falk RH Comenzo RL Skinner M The systemic amyloidosis NEJM 1997 337 898 909 9302305 10.1056/NEJM199709253371306
Cohen AS amyloidosis Bull Rheum Dis 1991 40 1 12 2059761
Grateau G The relation between Familial Mediterranean fever and amyloidosis Curr Opin Rheumatol 2000 12 61 64 10647956 10.1097/00002281-200001000-00010
Jensen LE Whitehead AS Regulation of serum amyloid A protein expression during the acute-phase response Biochem J 1998 334 489 503 9729453
Kerr R Interleukin 6 and hemostasis Br J Haematol 2001 115 3 12 11722403 10.1046/j.1365-2141.2001.03061.x
|
16138931
|
PMC1232862
|
CC BY
|
2021-01-04 16:38:37
|
no
|
Int Semin Surg Oncol. 2005 Sep 2; 2:17
|
utf-8
|
Int Semin Surg Oncol
| 2,005 |
10.1186/1477-7800-2-17
|
oa_comm
|
==== Front
J NeuroinflammationJournal of Neuroinflammation1742-2094BioMed Central London 1742-2094-2-201615689510.1186/1742-2094-2-20ResearchActivation of microglial NADPH oxidase is synergistic with glial iNOS expression in inducing neuronal death: a dual-key mechanism of inflammatory neurodegeneration Mander Palwinder [email protected] Guy C [email protected] Biochemistry Department, University of Cambridge, Tennis Court Road, Cambridge, CB2 1QW, UK2005 12 9 2005 2 20 20 25 7 2005 12 9 2005 Copyright © 2005 Mander and Brown; licensee BioMed Central Ltd.2005Mander and Brown; 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-activated glia are seen in many CNS pathologies and may kill neurons through the release of cytotoxic mediators, such as nitric oxide from inducible NO synthase (iNOS), and possibly superoxide from NADPH oxidase (NOX). We set out to determine the relative role of these species in inducing neuronal death, and to test the dual-key hypothesis that the production of both species simultaneously is required for significant neuronal death.
Methods
Primary co-cultures of cerebellar granule neurons and glia from rats were used to investigate the effect of NO (from iNOS, following lipopolysaccharide (LPS) and/or cytokine addition) or superoxide/hydrogen peroxide (from NOX, following phorbol 12-myristate 13-acetate (PMA), ATP analogue (BzATP), interleukin-1β (IL-1β) or arachidonic acid (AA) addition) on neuronal survival.
Results
Induction of glial iNOS caused little neuronal death. Similarly, activation of NOX alone resulted in little or no neuronal death. However, if NOX was activated (by PMA or BzATP) in the presence of iNOS (induced by LPS and interferon-γ) then substantial delayed neuronal death occurred over 48 hours, which was prevented by inhibitors of iNOS (1400W), NOX (apocynin) or a peroxynitrite decomposer (FeTPPS). Neurons and glia were also found to stain positive for nitrotyrosine (a putative marker of peroxynitrite) only when both iNOS and NOX were simultaneously active. If NOX was activated by weak stimulators (IL-1β, AA or the fibrillogenic prion peptide PrP106-126) in the presence of iNOS, it caused microglial proliferation and delayed neurodegeneration over 6 days, which was prevented by iNOS or NOX inhibitors, a peroxynitrite decomposer or a NMDA-receptor antagonist (MK-801).
Conclusion
These results suggest a dual-key mechanism, whereby glial iNOS or microglial NOX activation alone is relatively benign, but if activated simultaneously are synergistic in killing neurons, through generating peroxynitrite. This mechanism may mediate inflammatory neurodegeneration in response to cytokines, bacteria, ATP, arachidonate and pathological prions, in which case neurons may be protected by iNOS or NOX inhibitors, or scavengers of NO, superoxide or peroxynitrite.
microgliaperoxynitritenitric oxideprion proteininflammationcytokines
==== Body
Background
Glia (microglia and astrocytes) can become inflammation activated in many CNS pathologies, including infectious, ischaemic, inflammatory and neurodegenerative disorders [1,2]. Glial activation may be protective to the host, as it can lead to the removal of cell debris and killing of pathogens [3]. However excessive or chronic glial activation can kill nearby neurons [4,5]. Thus inflammation may contribute to many CNS pathologies including Alzheimer's, Parkinson's and motor neuron diseases, multiple sclerosis, meningitis, AIDS dementia, strokes, trauma and normal brain ageing [6,7]. It is therefore important to understand the mechanisms by which inflammatory-activated glia kill neurons.
Astrocytes and microglia can become activated by a range of factors, including pathogens and pro-inflammatory cytokines, and can lead to the subsequent death of co-cultured neurons [8,9]. Activated astrocytes and/or microglia produce a variety of factors which can mediate neuronal death, including reactive oxygen species (ROS) [10,11], nitric oxide [8,9,12] and glutamate [8,13], as well as pro-inflammatory cytokines that perpetuate glial activation, such as interleukin-1β (IL-1β) and tumour necrosis factor-α (TNF-α) [14].
The neuroprotective effects of anti-oxidants have been established [15] and are thought to be due to the removal of ROS (such as superoxide) and as well as more toxic molecules (such as peroxynitrite) [16]. There is evidence that NADPH oxidase is activated in Alzheimer's disease and AIDS dementia [17-19]. The major source of ROS during inflammation is NADPH oxidase [20,21], although other sources may also contribute [22,23]. NADPH oxidase is expressed mainly by microglia in the brain [21,24], and produces superoxide (O2-) extracellularly or within phagocytic vesicles, in order to kill pathogens. The oxidase can be acutely activated by PMA, ATP, arachidonic acid, some chemokines and cytokines [25-28]. Superoxide is then broken down mainly by extracellular and intracellular superoxide dismutase to give hydrogen peroxide (H2O2).
iNOS is not normally expressed in the brain, but is induced in astrocytes and microglia by proinflammatory cytokines and pathogen components, such as lipopolysaccharide (LPS)/endotoxin of Gram-negative bacteria [29]. Once expressed iNOS produces high, sustained levels of NO which can, in certain conditions, kill nearby neurons, by mechanisms including inhibition of mitochondrial respiration and the release of glutamate from neurons and glia, resulting in excitotoxicity [8]. However, such mechanisms may require a relatively high level of NO and/or a relatively low level of oxygen [30,31]. An alternative mechanism would be for NO to react with superoxide (e.g. from the NADPH oxidase) to produce peroxynitrite (ONOO-), which is potentially more neurotoxic to neurons than NO or superoxide [32,33].
This suggests a dual-key hypothesis of inflammatory neurodegeneration whereby iNOS expression or NADPH oxidase activation alone is relatively benign, but when combined together at the same time causes neurodegeneration via peroxynitrite. We have previously shown that acute activation of the NADPH oxidase in isolated microglia expressing iNOS results in the rapid disappearance of NO and produces ONOO- [32]. In this paper we report that activation of the microglial NADPH oxidase to produce superoxide is synergistic with NO from iNOS in inducing death of co-cultured neurons, whereas activation of either alone causes little or no death of co-cultured neurons.
Materials & methods
Materials
The following materials were purchased from the indicated sources: 1400W.dihydrochloride from Alexis (Nottingham, UK); MK-801 maleate, apocynin and FeTPPS (5,10,15,20-Tetrakis(4-sulfonatophenyl)porphyrinato Iron (III) chloride) from Calbiochem (Nottingham, UK). All other reagents were ordered from Sigma (Poole, UK).
Neuronal-glial culture
Cerebellar granule cell (CGC) cultures were prepared from 7-day-old Wistar rats, as described in Bal-Price & Brown, 2001. Briefly, the pups were anaesthetised using 5% halothane in oxygen, followed by decapitation. Brains were removed under sterile conditions and the cerebellum dissected. Meninges were removed and the cerebella dissociated in Versene solution (1:5000, Gibco BRL) and plated at 0.25 × 106 cells/cm2 in 24-well plates (in 500 μl DMEM) coated with 0.001% poly-L-lysine. Cultures were maintained in DMEM (Gibco BRL) supplemented with 5% horse serum, 5% foetal calf serum, 38 mM glucose, 5 mM HEPES, 2 mM glutamine, 25 mM KCl and 10 μg/ml gentamicin. Cells were kept at 37°C in a humidified atmosphere of 5% CO2/95% air and used for experiments at 16–18 days in vitro (DIV). Cultures of CGC's contained 22 ± 4% astrocytes and 2 ± 1% microglia as assessed by immunocytochemistry using antibodies against glial fibrillary acidic protein (GFAP: a marker for astrocytes) and complement receptor-3 (a marker for microglia), CGC's were identified based on morphology and at 16–18 DIV 76 ± 5% of the cells in the culture were CGC's. All experiments were undertaken in accordance with the UK Animals (Scientific Procedures) Act 1986.
Activation of glia in neuronal-glial cultures
Lipopolysaccharide (LPS), a cell wall component of Gram-negative bacteria and interferon-γ (IFN-γ), a pro-inflammatory cytokine, are potent activators of glia when administered together. Neuronal-glial cultures were treated with 100 ng/ml LPS (from Salmonella typhimurium) and 10 ng/ml IFN-γ (rat recombinant, Sigma) for 48 hours (or longer where indicated). The proinflammatory cytokines tumour necrosis factor-α (TNF-α; 10 ng/ml, rat recombinant, Sigma) and interleukin-1β (IL-1β; 10 ng/ml, rat recombinant, Sigma) were also used in combination with IFN-γ to activate the glia in neuronal-glial cultures (48 hours). Where present, inhibitors were added at the same time as LPS/IFN-γ.
In some experiments IL-1β or arachidonic acid (AA, 30 μM) were added to the cultures as well as LPS/IFN-γ. In these experiments, IL-1β or AA were added 24 hours after LPS/IFN-γ addition, but inhibitors were added at the same time as LPS/IFN-γ. Activators, inhibitors and IL-1β or AA were added once only and neuronal death was assessed 144 hours after LPS/IFN-γ addition.
In some experiments, prion protein or a fragment of the human prion protein were used (kindly provided by David R. Brown, University of Bath). Recombinant mouse prion protein was expressed in bacteria and purified using a histidine-tagging method, as described previously [34]. The prion peptide (PrP106-126) with sequence KTNMKHMAGAAAAGAVVGGLG was derived from amino acid residues 106–126 of the human prion protein sequence, and a scrambled sequence of the peptide was used as a control; sequence: NGAKALMGGHGATKVMVGAAA. Prion protein was used at 5 μg/ml and the prion protein peptides at 225 μg/ml.
To activate NADPH oxidase, phorbol 12-myristate 13-acetate (PMA, 50 ng/ml) or benzoyl(benzoyl)-ATP (BzATP, 1 mM) are used and are added to neuronal-glial cultures either alone or at the same time as LPS/IFN-γ.
Enrichment of microglia in neuronal-glial cultures
Primary rat microglia were obtained from mixed glial cultures (astrocytes and microglia). Glial cultures were prepared from the cerebral cortices of 7-day-old Wistar rats (the same brains that were used to isolate cerebellar granule neurons). Meninges were removed from the cerebral hemispheres and then dissociated using a solution of EBSS containing 0.3% BSA, 0.004% DNase I and 0.025% Trypsin. Cells were plated at 0.1 × 106 cells/cm2 in 75 cm2 cell culture flasks (Falcon) coated with 0.0005% poly-L-lysine. Cultures were maintained in DMEM supplemented with 10% foetal calf serum and 1% Penicillin-Streptomycin. Cells were kept at 37°C in a humidified atmosphere of 5% CO2/95% air.
At confluency, glial cultures were used to isolate microglial cells by gently shaking/tapping the mixed glial cultures to dislodge microglia loosely attached to astrocytes. Medium from the mixed glial cultures, containing microglia was removed and centrifuged (135 g for 5 minutes). Microglia were re-suspended in conditioned medium from CGC cultures and added to neuronal-glial cultures in some experiments (50, 000 microglia/cm2). Fifteen minutes after the addition of microglia to some neuronal-glial cultures, LPS/IFN-γ and inhibitors where appropriate were added together. Neuronal death was assessed 48 hours after LPS/IFN-γ addition.
Assessment of glial activation
Activation of glia in the neuronal-glial culture was assessed by NADPH diaphorase staining and measurements of nitrite in the medium. Nitric oxide synthase (NOS) is an NADPH diaphorase, using a chromogen (nitroblue tetrazolium, NBT), and NADPH as the reductant, diaphorase staining was used to detect cells with NOS activity. Following treatment (with cytokines or untreated for control staining) the neuronal glial cultures were fixed with 4% paraformaldehyde in phosphate buffer for 30 minutes at 4°C. After fixation, cells were incubated in 0.3% Triton X-100 (in phosphate buffer) for 5 minutes. Cells were then incubated for 2 hours at 37°C in 0.3% Triton X-100 containing 1 mg/ml NADPH and 0.2 mg/ml NBT. Cells were washed once with 0.3% Triton X-100 and then viewed using an inverted light microscope (Leica).
Nitrite levels in the medium were measured using the Griess reaction. Briefly, aliquots of medium following treatments were taken and centrifuged (8000 g for 5 minutes). 6 mM HCl was added to the supernatant and then 1 mM sulfanilamide and 1 mM N-1 (1-naphthyl)ethylenediamine (NEDA) were added. Absorbance at a wavelength of 548 nm was measured by plate reader (BMG, Fluostar Optima), before and after the addition of NEDA. Nitrite concentrations in samples were calculated from a standard curve of sodium nitrite in DMEM.
Assessment of cell viability
The viability of CGC's was assessed by propidium iodide (PI, 2 μg/ml) and Hoechst 33342 (6 μg/ml) staining, using a fluorescence microscope (Axiovert S-100) and filters for excitation at 365 nm and emission at 420 nm. The cell-impermeable nuclear dye, PI stains the nuclei of cells that have lost plasma membrane integrity and are considered to be necrotic. Using the cell-permeable DNA dye Hoechst 33342, the nuclear morphology of the CGC's was studied. Neuronal nuclei exhibiting irregular Hoechst staining, nuclear shrinkage, chromatin condensation and/or nuclear fragmentation but PI negative were classified as showing chromatin condensation (CC). Individual cells exhibiting both CC and PI staining were included in the PI data. Cells were counted in three microscopic fields in each well (3 wells per treatment) and expressed as a percentage of the total number of neurons. Each treatment was repeated at least three times.
Assessment of microglia proliferation
Microglia cells were identified using Isolectin IB4 (from Griffonia simplicifolia), which has strong affinity for microglia but not astrocytes. An Alexa Fluor 488 conjugate of isolectin IB4 (10 ng/ml) was added to cultures activated with LPS/IFN-γ and treated with IL-1β, AA or prion protein/peptide and incubated for 15 minutes at 37°C. Stained cells (microglia) were visualised and counted by viewing under a fluorescence microscope (excitation 488 nm, emission 530 nm).
3-nitrotyrosine immunocytochemistry
Mixed neuronal-glial cultures were stained for the peroxynitrite marker, 3-nitrotyrosine (3-NT). Cultures were untreated (control) or treated with LPS/IFN-γ, PMA, LPS/IFN-γ/PMA or FeTPPS + LPS/IFN-γ/PMA. Cultures were fixed with 4% paraformaldehyde and then incubated with 10 μg/ml of anti-nitrotyrosine monoclonal antibody (Upstate). The primary antibody was detected using a Cy3-conjugated secondary antibody (Jackson ImmunoResearch Laboratories). 3-NT -positive cells were visualised using a fluorescence microscope (excitation 546 nm, emission 590 nm).
Statistical analysis
Data are expressed as mean ± SEM and were analysed for significance using ANOVA.
Results
Inflammatory activation of glia in neuronal-glial cultures does not lead to substantial death of the co-cultured neurons
A mature mixed culture (16–18 days in vitro) of cerebellar granule neurons and glia (22% astrocytes and 2% microglia) was used to investigate inflammation-activated glia-induced neuronal death. The glia in the neuronal-glial cultures were activated with a combination of endotoxin (lipopolysaccharide, LPS) and a pro-inflammatory cytokine (interferon-γ, IFN-γ) or different combinations of pro-inflammatory cytokines including tumour necrosis factor-α (TNF-α) and interleukin-1β (IL-1β). Neuronal death was assessed 48 hours after treatment with the inflammatory activators (LPS/IFN-γ, TNF-α/IFN-γ, IL-1β/IFN-γ or TNF-α/IL-1β/IFN-γ). Two nuclear dyes were used to stain the cultures and assess for necrosis and apoptosis: cell-impermeable propidium iodide (PI) to stain necrotic cells and the cell-permeable Hoechst 33342 used to characterise any neuronal nuclei showing signs of chromatin condensation or nuclear fragmentation (characteristic of apoptosis). Although relatively small, significant levels of neuronal death were induced following activation with LPS/IFN-γ, TNF-α/IFN-γ or TNF-α/IL-1β/IFN-γ but not IL-1β/IFN-γ (Table 1).
Table 1 Effects of inflammatory activated-glia in mixed neuronal-glial cultures on neuronal death. Neuronal death was assessed by propidium iodide staining (PI, necrosis) or chromatin condensation of neuronal nuclei by Hoechst 33342 staining (CC, a marker of apoptosis) 48 hours after treatment. Nitrite (the primary breakdown product of NO) levels were measured in the culture medium 48 hours following treatments. Statistical differences were established using ANOVA at *p < 0.05 and ***p < 0.001. Data expressed is mean ± SEM, n = 3 or more.
TREATMENT PI (%) CC (%) NITRITE (μM)
UNTREATED 0.9 ± 0.9 0.5 ± 0.6 2.7 ± 3.0
LPS/IFN-γ 5.7 ± 3.4 * 3.6 ± 1.5 * 18.6 ± 8.4 ***
TNF-α/IFN-γ 5.6 ± 0.4 *** 4.3 ± 2.9 * 4.2 ± 2.8
IL-1β/IFN-γ 1.1 ± 1.1 0.6 ± 0.7 3.7 ± 2.1
TNF-α/IL-1β/IFN-γ 6.3 ± 4.1 * 5.5 ± 3.2 * 4.5 ± 1.4
To confirm that the glia in the culture had actually been activated to express iNOS, we used a simple stain for nitric oxide synthase (NOS) activity (NADPH diaphorase staining) enabling us to visualise cells with NOS activity and distinguish between microglia and astrocytes based on morphology. Additionally we assessed nitrite levels in the culture medium as a measure of NO production (Table: 1). Non-activated cultures showed no NADPH diaphorase staining in glia, but low-level staining was seen in neurons (probably due to nNOS) and correlates with the low level of nitrite present in the medium (Figure: 1a; Table: 1). However, after treatment with LPS/IFN-γ, a high proportion of glia (both microglia and astrocytes) stained intensely for diaphorase activity (Figure: 1b). Treatment with TNF-α/IFN-γ, IL-1β/IFN-γ or TNF-α/IL-1β/IFN-γ resulted in much less diaphorase staining of glia, and little or no nitrite elevation, indicating a requirement for LPS to induce substantial iNOS expression.
Figure 1 NOS activity in mature neuronal-glial cultures. NADPH diaphorase staining was used to assess for NOS activity. Non-activated (control) cultures show weak staining in neurons and along their processes (a), but following LPS/IFN-γ treatment (b) dark staining is visible in glia (astrocytes and microglia). The photographs shown are representative and were taken 48 hours after treatment, n>3.
Relatively pure neuronal cultures (CGC cultures isolated as described in the methods section and then treated with 10 μM arabinoside cytosine at 18 hours to inhibit the proliferation of glia) consisting of 97 ± 4% neurons, 2 ± 1% astrocytes and 1 ± 1% microglia were not affected by the presence of cytokines alone (mean % of chromatin-condensed (CC) and propidium iodide-positive (PI) neurons ± SEM of 3 separate cultures; control: CC: 4 ± 3%, PI: 8 ± 4%; 10 ng/ml IL-1β: CC: 3 ± 2%, PI: 7 ± 3%; 10 ng/ml TNF-α: CC: 3 ± 2%, PI: 9 ± 4%). Additionally, no significant adverse effects were seen even if the concentrations of IL-1β or TNF-α were increased 10-fold (mean % of neurons ± SEM of 1 culture; control: CC: 4 ± 3%, PI: 8 ± 4%; 100 ng/ml IL-1β: CC: 4 ± 2%, PI: 5 ± 4%; 100 ng/ml TNF-α: CC: 4 ± 2%, PI: 6 ± 4%) or if combined with 10 ng/ml IFN-γ treatment (mean % of neurons ± SEM of 2 separate cultures; control: CC: 4 ± 3%, PI: 8 ± 4%; 10 ng/ml IL-1β + IFN-γ: CC: 3 ± 2%, PI: 5 ± 3%; 10 ng/ml TNF-α + IFN-γ: CC: 3 ± 3%, PI: 7 ± 2%).
These results suggest that the cytokines have no direct toxicity for neurons, and although nitric oxide (NO) is produced by iNOS expressed in glia following activation with LPS/IFN-γ, it is not able to kill the co-cultured neurons alone, or the quantities of NO produced are not sufficient to induce widespread death of these mature neuronal cultures.
Simultaneous activation of iNOS and NADPH oxidase results in massive neuronal death, mediated by peroxynitrite
As NO produced by inflammatory activated glia did not induce substantial neuronal death, we investigated whether simultaneous production of superoxide resulting in peroxynitrite would be more toxic to neurons. Peroxynitrite is formed from the diffusion-limited reaction of NO with superoxide. Under inflammatory conditions in the brain, NADPH oxidase is the major source of superoxide, therefore we used phorbol 12-myristate 13-acetate (PMA) to activate this enzyme and generate a source of superoxide in the neuronal-glial culture. As the number of NADPH diaphorase-positive glia was greatest following treatment with LPS/IFN-γ, we used LPS/IFN-γ to induce iNOS expression in the glia and provide a source of NO. We found that treating neuronal-glial cultures with LPS/IFN-γ/PMA for 48 hours induced extensive neuronal death (Figure: 2a). Treatment of the cultures with PMA alone induced only low levels of neuronal death, similar to that seen with LPS/IFN-γ treatment alone. However, activation of both NADPH oxidase and iNOS was synergistic in inducing neuronal death. This neuronal death was prevented by an iNOS inhibitor of (1400W), a NADPH oxidase (apocynin) a peroxynitrite scavenger (FeTPPS), but not by a blocker of the NMDA receptor (MK-801).
Figure 2 Activation of NADPH oxidase in the presence of glial iNOS is synergistic in killing co-cultured neurons. Cultures were stained with the cell-impermeable dye propidium iodide (PI) to count necrotic cells and the cell-permeable dye Hoechst 33342 to count neuronal nuclei showing chromatin condensation/fragmentation (CC), 48 hours after treatment (a). PMA stimulation of NADPH oxidase did not substantially affect neuronal survival, but in the presence of LPS/IFN-γ had synergistic effects on neuronal death, which were blocked by inhibitors of iNOS (25 μM 1400W), NADPH oxidase (1 mM apocynin), or a peroxynitrite scavenger (100 μM FeTPPS) but not by a blocker of the NMDA receptor (10 μM MK-801). Nitrite and nitrate levels were not affected by the presence of PMA or apocynin but were significantly reduced by 1400W (b). Statistical differences were established using ANOVA at *p <0.05, **p < 0.01 and ***p < 0.001, the symbol # replaces * when comparing protection against LPS/IFN-γ/PMA induced neuronal death. The symbol ¶ is used to demonstrate a significant difference in comparison to PMA or LPS/IFN-γ alone. Statistical significance refers to the total death (black + white parts of the bar). Data expressed is mean ± SEM, n = 3 or more.
As PMA activates the protein kinase C pathway, the effects of PMA might be due to reasons other than stimulating the microglial NADPH oxidase, such as increased iNOS expression leading to more NO production and neuronal death by NO and not peroxynitrite. However, the levels of nitrite and nitrate in the culture medium of neuronal-glial cultures treated with LPS/IFN-γ/PMA were not different to those found in the absence of PMA (Figure: 2b).
To determine whether peroxynitrite generated by glia reaches the neurons, the neuronal-glial cultures were tested for nitrotyrosine immunoreactivity. Positively stained neurons (and glia) were only seen following treatment with LPS/IFN-γ/PMA (Figure: 3) and not in the presence of the peroxynitrite scavenger FeTPPS or when treated with LPS/IFN-γ (data not shown) or PMA alone (data not shown). However, no PI-positive glia or changes in glial nuclear morphology were observed in any of the conditions, implying that although they were exposed to peroxynitrite it did not induce glial death.
Figure 3 Activation of NADPH oxidase in the presence of iNOS expression leads to 3-nitrotyrosine immunoreactivity in neurons and glia. Neuronal-glial cultures treated with LPS/IFN-γ/PMA for 48 hours showed extensive immunoreactivity for 3-nitrotyrosine, which was absent in the presence of FeTPPS. Untreated cultures (control) showed no staining for 3-nitrotyrosine. The photographs shown are representative and were taken 48 hours after treatment, n>3.
ATP is known to be released by neurons and glia in a variety of conditions, and has been reported to activate the microglial NADPH oxidase via P2X7 receptors [26]. We found that ATP rapidly stimulated superoxide/hydrogen peroxide production by isolated microglia, which was sensitive to diphenyleneiodonium (DPI), an inhibitor of NADPH oxidase (ATP: 80 ± 7 picomoles H2O2/minute/1 × 105 microglia). However, ATP did not induce neuronal death alone, or in synergy with LPS/IFN-γ treatment (data not shown), probably because it is rapidly hydrolysed in cell culture medium [35]. Therefore, we used a non-hydrolysable ATP analogue, 2'-3'-O-(4- benzoylbenzoyl)-ATP (BzATP), known to be a specific P2X7 receptor agonist [36]. BzATP was also found to stimulate DPI-sensitive hydrogen peroxide production by isolated microglia, which was comparable to that produced by PMA (control: 12 ± 3; PMA: 204 ± 50; BzATP: 124 ± 15 picomoles H2O2/minute/1 × 105 microglia). BzATP did not induce neuronal death alone but had synergistic effects on neuronal death in the presence iNOS expression (Figure: 4). LPS/IFN-γ/BzATP induced neuronal death was blocked by inhibitors of iNOS, NADPH oxidase and a peroxynitrite scavenger, but not by the NMDA receptor blocker.
Figure 4 NADPH oxidase stimulation by P2X7 receptor activation in the presence of glial iNOS kills co-cultured neurons. Neuronal death was assessed by propidium iodide staining (PI) and chromatin condensation of neuronal nuclei by Hoechst 33342 staining (CC) 48 hours after treatment. Neuronal death induced by BzATP following LPS/IFN-γ activation, was prevented by inhibitors of iNOS (25 μM 1400W), NADPH oxidase (1 mM apocynin) and a peroxynitrite scavenger (100 μM FeTPPS) but not by a blocker of the NMDA receptor (10 μM MK-801). Statistical differences were established using ANOVA at *p < 0.05 and ***p < 0.001, the symbol # replaces * when comparing protection against LPS/IFN-γ/BzATP induced neuronal death. Statistical significance refers to the total death (black + white parts of the bar). Data expressed is mean ± SEM, n = 3 or more.
Activation of glia in microglia-enriched neuronal-glial cultures potently kills co-cultured neurons
We have found that IL-1β or arachidonic acid (AA) can activate the microglial NADPH oxidase, although to lesser extent than PMA (control: 12 ± 3; IL-1β: 37 ± 20; AA: 24 ± 4 picomoles H2O2/minute/1 × 105 microglia). We therefore tested whether IL-1β or AA could synergise with LPS/IFN-γ to induce neuronal death. The addition of either IL-1β or AA did not induce further neuronal death than that induced by LPS/IFN-γ alone up to 48 hours after additions (data not shown). However if such cultures were maintained for 6 days, we found that widespread neuronal death occurred (Figure: 5a, b) and was blocked by inhibitors of iNOS, NADPH oxidase, a peroxynitrite scavenger and a blocker of the NMDA receptor. Treatment with IL-1β or AA alone did not have any effect on neuronal survival, but did increase the number of microglia in neuronal-glial cultures (Figure: 5c). Treatment with LPS/IFN-γ was found to inhibit microglia proliferation but in the presence of IL-1β or AA this inhibition was overcome and lead to a progressive increase in the number of microglia and subsequent neuronal death. The mitogenic effects of IL-1β or AA are probably mediated by hydrogen peroxide following stimulation of NADPH oxidase (unpublished data) and we found that the NADPH oxidase inhibitor, apocynin, prevented this increase in the number of microglia. Nitrite and nitrate (NOX) levels (Figure: 5d) were higher in cultures treated with IL-1β or AA plus LPS/IFN-γ, but not in the presence of apocynin, which blocked microglial proliferation, suggesting that microglia were the predominant source of NO and/or peroxynitrite.
Figure 5 Effects of IL-1β or arachidonic acid (AA) on neuronal survival in the presence of inflammation-activated glia in neuronal-glial cultures. Neuronal death was assessed by propidium iodide staining (PI; a) or chromatin condensation of neuronal nuclei (CC; b) after 6 days of treatment. Neuronal death was prevented by inhibitors of iNOS (25 μM 1400W), NADPH oxidase (1 mM apocynin), a blocker of the NMDA-receptor (10 μM MK-801) or a peroxynitrite scavenger (100 μM FeTPPS). Neuronal death was accompanied by proliferation of microglia (c). Microglial proliferation was inhibited by LPS/IFN-γ treatment alone but in the presence of IL-1β or AA it was stimulated and returned to basal levels. This stimulation of proliferation by IL-1β or AA (in the presence of LPS/IFN-γ) was completely prevented by apocynin. Additionally, nitrite/nitrate (NOX) levels correlated with the number of microglia present (d). Statistical differences were established using ANOVA at *p < 0.05, **p < 0.01 and ***p < 0.001, the symbol * is used when assessing prevention of neuronal death in comparison to LPS/IFN-γ with IL-1β or AA. The symbol ¶ is used when comparing neuronal death to that induced by LPS/IFN-γ alone and # when comparing neuronal death induced by IL-1β or AA treatment alone. In c & d, the differences are in comparison to IL-1β or AA alone (*), LPS/IFN-γ (¶) or LPS/IFN-γ plus IL-1β or AA (#). Data expressed is mean ± SEM, n = 3 or more.
Since IL-1β and AA stimulated microglial proliferation (in the presence or absence of LPS/IFN-γ), we wanted to test whether increasing the microglial density would sensitise to LPS/IFN-γ induced neuronal death. So we investigated whether enriching the microglia population in the neuronal-glial culture followed by inflammatory activation would result in widespread neuronal death. The neuronal-glial culture used in the last section was enriched with microglia by adding freshly isolated microglia. LPS/IFN-γ activation of a microglia-rich (15% microglia as opposed to 2%) neuronal-glial culture resulted in all neurons rapidly losing their dendritic processes and shrinkage of the cell body (Figure: 6b), in addition to chromatin condensation or propidium iodide staining of the nuclei at 48 hours of treatment (Figure: 6a). This neuronal death was prevented by inhibitors of iNOS, NADPH oxidase, a peroxynitrite decomposition catalyst and a blocker of the NMDA receptor. The addition of microglia alone (non-activated) did not affect neuronal survival (Figure: 6a, untreated). In support of microglia as the key cell type in inflammatory neurodegeneration, LPS/IFN-γ activation of astrocyte-enriched neuronal-glial cultures did not lead to widespread killing of co-cultured neurons (isolation of neuronal-glial cultures as normal but plated onto a confluent bed of astrocytes; data not shown).
Figure 6 Activation of microglia-enriched neuronal-glial cultures induces complete neurodegeneration. The microglia population was enriched in neuronal-glial cultures by adding isolated microglia (50,000 microglia/cm2). Neuronal death was assessed by propidium iodide staining (PI) or chromatin condensation of neuronal nuclei (CC) at 48 hours after treatment (a). Neuronal death was prevented by inhibitors of iNOS (25 μM 1400W), NADPH oxidase (1 mM apocynin), a blocker of the NMDA-receptor (10 μM MK-801), or a peroxynitrite scavenger (100 μM FeTPPS). LPS/IFN-γ activation of the microglia-enriched neuronal-glial cultures led to complete disintegration of neuronal processes and severe shrinkage of neuronal cell bodies (b). Statistical differences were established using ANOVA at *p < 0.05 and ***p < 0.001, in comparison to control (added microglia) non-activated cultures, and the symbol # replaces * when comparing protection against neuronal death induced by LPS/IFN-γ activated cultures. Statistical significance refers to the total death (black + white parts of the bar). Data expressed is mean ± SEM, n = 3 or more. Photographs shown are representative and were taken 48 hours after the addition of LPS/IFN-γ.
Prion protein or PrP106-126 induce neuronal death in the presence of inflammatory activation mediated by microglia and NADPH oxidase activation
The prion peptide, PrP106-126, has previously been shown to activate microglia, causing proliferation and ROS production [37,38]. We have recently found that the prion protein and peptide stimulate the NADPH oxidase in isolated microglia (control: 12 ± 3; prion protein: 29 ± 3; PrP106-126: 38 ± 13 picomoles H2O2/minute/1 × 105 microglia). We decided to investigate whether the addition of PrP106-126 to iNOS-expressing glia in neuronal-glial cultures would also lead to delayed neurodegeneration, mediated by peroxynitrite and microglia. The addition of prion protein or PrP106-126 alone did not affect neuronal survival in these mature neuronal-glial cultures, but did lead to microglial proliferation (Table: 2). In the presence of glial iNOS (following LPS/IFN-γ treatment), PrP106-126 or prion protein did not exacerbate neuronal death over a period of 2 days, but were synergistic in killing the co-cultured neurons at 6 days (Figure: 7a, b), while a scrambled peptide of the PrP106-126 sequence had no effect. Neuronal death was prevented by blocking NO production from iNOS (1400W), or superoxide from NADPH oxidase (apocynin), through the removal of peroxynitrite (FeTPPS), or by inhibiting the NMDA receptor (MK-801). Additionally, neuronal death was accompanied by microglia proliferation, which was blocked by apocynin (Figure: 7c). Nitrite/nitrate levels were also suppressed in the presence of apocynin, as well as 1400W (Figure: 7d).
Table 2 Prion protein or peptide (PrP106-126) does not affect neuronal survival. Neuronal-glial cultures treated once with either prion protein (5 μg/ml) or PrP106-126 (225 μg/ml) did not induce neuronal death over a period of 7 days (assessed by Hoechst 33342 to visualise chromatin condensation (CC) or propidium iodide (PI) to stain necrotic cells). However, prion protein or PrP106-126 did stimulate the proliferation of microglia in neuronal-glial cultures over the same period of time. Statistical differences were established using ANOVA at *p < 0.05, **p < 0.01 and ***p < 0.001 and are in comparison to untreated cultures (symbol *); data expressed is mean ± SEM, n = 3 or more.
Treatment PI (%) CC (%) Microglia per field
Untreated 2.0 ± 1.6 0.6 ± 0.3 22 ± 5
Prion protein 2.9 ± 0.5 0.7 ± 0.6 53 ± 8 ***
PrP106-126 1.0 ± 0.8 0.7 ± 0.4 51 ± 7 ***
Figure 7 Delayed neurodegeneration induced by prion protein or PrP106-126 in the presence of iNOS expression is microglia-dependent and mediated by peroxynitrite. The addition of prion protein (5 μg/ml) or PrP106-126 (225 μg/ml) to LPS/IFN-γ treated neuronal-glial cultures induced delayed death of co-cultured neurons, over 6 days. Neuronal death, assessed by Hoechst 33342 to visualise chromatin condensation (CC; b) or PI for necrosis (a) was prevented by inhibitors of iNOS (25 μM 1400W) and NADPH oxidase (1 mM apocynin), a peroxynitrite scavenger (100 μM FeTPPS) or a blocker of the NMDA receptor (10 μM MK-801). Neuronal death was accompanied by proliferation of microglia (c). Microglial proliferation was inhibited by LPS/IFN-γ treatment alone but in the presence of prion protein or PrP106-126 it was stimulated and returned to basal levels. This stimulation of proliferation by prion protein or PrP106-126 (in the presence of LPS/IFN-γ) was completely prevented by apocynin. Additionally, nitrite/nitrate (NOX) levels correlated with the number of microglia present (d). Statistical differences were established using ANOVA at *p < 0.05, **p < 0.01 and ***p < 0.001 are in comparison to untreated cultures (symbol *) or LPS/IFN-γ treatment (symbol ¶) or LPS/IFN-γ plus prion protein or PrP106-126 (symbol #); data expressed is mean ± SEM, n = 3 or more. In c & d, the differences are in comparison to prion protein or PrP106-126 alone (*), LPS/IFN-γ (¶) or LPS/IFN-γ plus prion protein or PrP106-126 (#). Data expressed is mean ± SEM, n = 3 or more.
Discussion
We found that LPS/IFN-γ induced NOS activity within cultured glia, but induced relatively little death of co-cultured neurons. It has previously been reported that LPS/cytokine-induced iNOS expression in glia is sufficient [5,8,39,40] or insufficient [41-43] to induce death of co-cultured neurons. Similarly, in vivo it has been reported that iNOS expression is sufficient [44,45] or insufficient [46-48] to induce neuronal death. This suggests that either there is a threshold level for NO/iNOS induced neuronal death [49], or NO/iNOS-induced neuronal death is conditional upon some other factors. We have recently reported one such conditional factor (hypoxia) that synergises with NO/iNOS to induce neuronal death [31]. In this report we have tested the hypothesis that NO/iNOS induced neuronal death is conditional upon microglial NADPH oxidase activation.
It has previously been shown that PMA stimulation of microglia results in superoxide production through stimulation of NADPH oxidase [50] and, in the presence of LPS/IFN-γ activated glia (producing NO from iNOS), the superoxide combines with NO to form peroxynitrite [32]. We found that if the NADPH oxidase was stimulated by PMA in the presence of LPS/IFN-γ activated glia, it resulted in extensive death of the co-cultured neurons, while PMA alone induced very little neuronal death. In pathophysiological conditions, extracellular levels of ATP can increase [51], and ATP can activate purinergic receptors (more specifically P2X7 receptors), which can lead to the activation of NADPH oxidase [26]. We used a specific P2X7 receptor agonist (BzATP) to activate the NADPH oxidase in the presence of iNOS expression (LPS/IFN-γ activated cultures) and we found extensive neuronal death, comparable to that induced by LPS/IFN-γ/PMA. Neuronal-glial cultures activated with LPS/IFN-γ/PMA or LPS/IFN-γ/BzATP induced delayed neuronal death that occurred over 2 days. This is partly due to the time taken for iNOS expression, but it also implies that once peroxynitrite is generated neuronal death is not immediate.
In both cases (LPS/IFN-γ/PMA or LPS/IFN-γ/BzATP), inhibitors of iNOS or NADPH oxidase or a scavenger of peroxynitrite prevented this neuronal death, implicating peroxynitrite as the potential mediator of neuronal death and the source of peroxynitrite as NO from iNOS and superoxide from NADPH oxidase. The putative peroxynitrite marker nitrotyrosine, was found in both neurons and some glia, implying that LPS/IFN-γ/PMA treatment does result in peroxynitrite production that reacts with neurons. Furthermore, the peroxynitrite decomposition catalyst prevents the occurrence of nitrotyrosine-positive neurons following LPS/IFN-γ/PMA treatment. FeTPPS has been shown to rapidly react and catalyse the decomposition of extracellular peroxynitrite [52] and inhibit tyrosine nitration [53]. The presence of nitrotyrosine immunoreactivity in glia did not appear to induce glial death. It has been found that glia can up-regulate their antioxidant defences to become more resistant to oxidative stress [54], which may explain the lack of change in glial morphology.
The mechanism of peroxynitrite-induced neuronal death is still unclear but has been proposed to involve DNA-damage induced PARP activation [55], damage to the mitochondrial respiratory chain [56], and lipid peroxidation [57]. It is still controversial whether peroxynitrite-induced neuronal death involves activation of the NMDA receptor [58,59]. We found that a blocker of the NMDA-receptor did not prevent the relatively acute neuronal death induced by LPS/IFN-γ/PMA or LPS/IFN-γ/BzATP, but did prevent the relatively slow neuronal death induced by LPS/IFN-γ/IL-1β or LPS/IFN-γ/AA, although in both cases death was prevented by a peroxynitrite decomposer. It is possible that low, sustained levels of peroxynitrite induce neuronal death via the NMDA receptor, whereas high, acute levels induce death by other means, but we have not directly tested this. We found that IL-1β or AA activated NADPH oxidase hydrogen peroxide production to a lesser extent than PMA but, like PMA, either IL-1β or AA synergised with LPS/IFN-γ to induce neuronal death mediated by peroxynitrite following activation of iNOS and NADPH oxidase. However the neuronal death induced by LPS/IFN-γ/IL-1β or LPS/IFN-γ/AA occurred over 6 days, rather than 2 days as with LPS/IFN-γ/PMA or LPS/IFN-γ/BzATP. This relative delay might be due to the lower level of NADPH oxidase activation and thus peroxynitrite production. Additionally, Il-1β or AA caused microglial proliferation during the 6-day cultures, which may have contributed to the delayed neuronal death. Recently we found that IL-1β or AA stimulated microglial proliferation in microglia-astrocyte cultures via hydrogen peroxide production from NADPH oxidase (manuscript in preparation). Here we have shown that IL-1β or AA stimulate the proliferation of microglia in neuronal-glial cultures, even in the presence of LPS/IFN-γ (which itself inhibits microglial proliferation). In order to test whether an increase in microglia would potentiate LPS/IFN-γ induced neuronal death, we added extra isolated microglia to the neuronal-glial culture, increasing the microglial population from 2% to 15% of cells in the co-culture. In such microglia-enriched cultures, LPS/IFN-γ induced neuronal death was greatly increased. These observations suggest that microglia are essential for inflammatory activated glia-induced neuronal death, and one reason for this may be the expression of NADPH oxidase, which is predominantly localised to microglia [24].
Transmissible spongiform encephalopathies (Prion diseases) are lethal neurodegenerative disorders characterised by the progressive accumulation of a protease resistant isoform (PrPsc) of the normal host prion protein (PrPc), in amyloid plaques [60]. An inflammatory response, predominantly mediated by microglia, is seen in post-mortem brain tissue, in transgenic models of the disease, and in culture [61]. A peptide, consisting of residues 106–126 (PrP106-126) of the human prion protein, replicates many of the pathological mechanisms involved in prion diseases and provides a good in vitro model. Contrary to published data [38,62], we observed no neurotoxicity following the addition of PrP106-126 alone to this mature neuronal-glial culture. We found that both prion protein and PrP106-126, but not a scrambled peptide, stimulated microglial proliferation when added to neuronal-glial cultures, and this proliferation was blocked by a NADPH oxidase inhibitor. Both prion protein and peptide were synergistic in killing neurons in the presence of glial iNOS, in a peroxynitrite and microglia-dependent mechanism. The addition of the cellular isoform of prion protein to cell cultures has previously been shown to have no toxic effects [34]. However, in the presence of glial iNOS we found it to induce significant levels of neuronal death, although significantly less than that induced by PrP106-126.
Conclusion
We have shown that in a mature mixed culture of neurons and glia, activation of iNOS or NADPH oxidase alone does not result in substantial neuronal death, but that simultaneous activation of both is synergistic in killing co-cultured neurons. This neuronal death appears to be dependent on microglia, and microglial proliferation is itself stimulated by activating the NADPH oxidase. These results suggest a dual-key hypothesis for inflammatory neurodegeneration; i.e. that activation of glial iNOS or NADPH oxidase alone may be relatively benign, but when activated together they cause peroxynitrite-mediated neuronal death. The conditionality of NO/iNOS-induced neuronal death provides insight into the mechanisms of inflammatory neurodegeneration and suggests that microglial NADPH oxidase may be a key therapeutic target.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
PM participated in the design of this study, did the lab work, data analysis and wrote major parts of the paper. GCB conceived the study, participated in its design and helped to draft the manuscript. Both authors read and approved the final manuscript.
Acknowledgements
This work was supported by the BBSRC, MRC and Alzheimer's Research Trust.
==== Refs
Eddleston M Mucke L Molecular profile of reactive astrocytes--implications for their role in neurologic disease Neuroscience 1993 54 15 36 8515840 10.1016/0306-4522(93)90380-X
Kreutzberg GW Microglia: a sensor for pathological events in the CNS Trends Neurosci 1996 19 312 318 8843599 10.1016/0166-2236(96)10049-7
Polazzi E Contestabile A Reciprocal interactions between microglia and neurons: from survival to neuropathology Rev Neurosci 2002 13 221 242 12405226
Banati RB Gehrmann J Schubert P Kreutzberg GW Cytotoxicity of microglia Glia 1993 7 111 118 8423058 10.1002/glia.440070117
Boje KM Arora PK Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death Brain Res 1992 587 250 256 1381982 10.1016/0006-8993(92)91004-X
Brown GC Bal-Price A Inflammatory neurodegeneration mediated by nitric oxide, glutamate, and mitochondria Mol Neurobiol 2003 27 325 355 12845153 10.1385/MN:27:3:325
McGeer PL McGeer EG The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodegenerative diseases Brain Res Brain Res Rev 1995 21 195 218 8866675 10.1016/0165-0173(95)00011-9
Bal-Price A Brown GC Inflammatory neurodegeneration mediated by nitric oxide from activated glia-inhibiting neuronal respiration, causing glutamate release and excitotoxicity J Neurosci 2001 21 6480 6491 11517237
Chao CC Hu S Sheng WS Bu D Bukrinsky MI Peterson PK Cytokine-stimulated astrocytes damage human neurons via a nitric oxide mechanism Glia 1996 16 276 284 8833198 10.1002/(SICI)1098-1136(199603)16:3<276::AID-GLIA10>3.0.CO;2-X
Beckman JS Chen J Crow JP Ye YZ Reactions of nitric oxide, superoxide and peroxynitrite with superoxide dismutase in neurodegeneration Prog Brain Res 1994 103 371 380 7886219
Chao CC Hu S Peterson PK Modulation of human microglial cell superoxide production by cytokines J Leukoc Biol 1995 58 65 70 7616108
Bolanos JP Almeida A Stewart V Peuchen S Land JM Clark JB Heales SJ Nitric oxide-mediated mitochondrial damage in the brain: mechanisms and implications for neurodegenerative diseases J Neurochem 1997 68 2227 2240 9166714
Barger SW Basile AS Activation of microglia by secreted amyloid precursor protein evokes release of glutamate by cystine exchange and attenuates synaptic function J Neurochem 2001 76 846 854 11158256 10.1046/j.1471-4159.2001.00075.x
Chao CC Hu S Ehrlich L Peterson PK Interleukin-1 and tumor necrosis factor-alpha synergistically mediate neurotoxicity: involvement of nitric oxide and of N-methyl-D-aspartate receptors Brain Behav Immun 1995 9 355 365 8903852 10.1006/brbi.1995.1033
Floyd RA Antioxidants, oxidative stress, and degenerative neurological disorders Proc Soc Exp Biol Med 1999 222 236 245 10601882 10.1046/j.1525-1373.1999.d01-140.x
Metodiewa D Koska C Reactive oxygen species and reactive nitrogen species: relevance to cyto(neuro)toxic events and neurologic disorders. An overview Neurotox Res 2000 1 197 233 12835102
Zekry D Epperson TK Krause KH A role for NOX NADPH oxidases in Alzheimer's disease and other types of dementia? IUBMB Life 2003 55 307 313 12938732
Shimohama S Tanino H Kawakami N Okamura N Kodama H Yamaguchi T Hayakawa T Nunomura A Chiba S Perry G Smith MA Fujimoto S Activation of NADPH oxidase in Alzheimer's disease brains Biochem Biophys Res Commun 2000 273 5 9 10873554 10.1006/bbrc.2000.2897
Boven LA Gomes L Hery C Gray F Verhoef J Portegies P Tardieu M Nottet HS Increased peroxynitrite activity in AIDS dementia complex: implications for the neuropathogenesis of HIV-1 infection J Immunol 1999 162 4319 4327 10201964
Qin L Liu Y Wang T Wei SJ Block ML Wilson B Liu B Hong JS NADPH oxidase mediates lipopolysaccharide-induced neurotoxicity and proinflammatory gene expression in activated microglia J Biol Chem 2004 279 1415 1421 14578353 10.1074/jbc.M307657200
Green SP Cairns B Rae J Errett-Baroncini C Hongo JA Erickson RW Curnutte JT Induction of gp91-phox, a component of the phagocyte NADPH oxidase, in microglial cells during central nervous system inflammation J Cereb Blood Flow Metab 2001 21 374 384 11323523 10.1097/00004647-200104000-00006
Sastre J Pallardo FV Vina J The role of mitochondrial oxidative stress in aging Free Radic Biol Med 2003 35 1 8 12826250 10.1016/S0891-5849(03)00184-9
Dugan LL Sensi SL Canzoniero LM Handran SD Rothman SM Lin TS Goldberg MP Choi DW Mitochondrial production of reactive oxygen species in cortical neurons following exposure to N-methyl-D-aspartate J Neurosci 1995 15 6377 6388 7472402
Klegeris A McGeer PL Rat brain microglia and peritoneal macrophages show similar responses to respiratory burst stimulants J Neuroimmunol 1994 53 83 90 8051299 10.1016/0165-5728(94)90067-1
Benna JE Dang PM Gaudry M Fay M Morel F Hakim J Gougerot-Pocidalo MA Phosphorylation of the respiratory burst oxidase subunit p67(phox) during human neutrophil activation. Regulation by protein kinase C-dependent and independent pathways J Biol Chem 1997 272 17204 17208 9202043 10.1074/jbc.272.27.17204
Parvathenani LK Tertyshnikova S Greco CR Roberts SB Robertson B Posmantur R P2X7 mediates superoxide production in primary microglia and is up-regulated in a transgenic mouse model of Alzheimer's disease J Biol Chem 2003 278 13309 13317 12551918 10.1074/jbc.M209478200
Shiose A Sumimoto H Arachidonic acid and phosphorylation synergistically induce a conformational change of p47phox to activate the phagocyte NADPH oxidase J Biol Chem 2000 275 13793 13801 10788501 10.1074/jbc.275.18.13793
Babior BM NADPH oxidase Curr Opin Immunol 2004 16 42 47 14734109 10.1016/j.coi.2003.12.001
Murphy S Simmons ML Agullo L Garcia A Feinstein DL Galea E Reis DJ Minc-Golomb D Schwartz JP Synthesis of nitric oxide in CNS glial cells Trends Neurosci 1993 16 323 328 7691008 10.1016/0166-2236(93)90109-Y
Bonfoco E Krainc D Ankarcrona M Nicotera P Lipton SA Apoptosis and necrosis: two distinct events induced, respectively, by mild and intense insults with N-methyl-D-aspartate or nitric oxide/superoxide in cortical cell cultures Proc Natl Acad Sci U S A 1995 92 7162 7166 7638161
Mander P Borutaite V Moncada S Brown GC Nitric oxide from inflammatory-activated glia synergizes with hypoxia to induce neuronal death J Neurosci Res 2005 79 208 215 15558752 10.1002/jnr.20285
Bal-Price A Matthias A Brown GC Stimulation of the NADPH oxidase in activated rat microglia removes nitric oxide but induces peroxynitrite production J Neurochem 2002 80 73 80 11796745 10.1046/j.0022-3042.2001.00675.x
Beckman JS Crow JP Pathological implications of nitric oxide, superoxide and peroxynitrite formation Biochem Soc Trans 1993 21 330 334 8395426
Brown DR Wong BS Hafiz F Clive C Haswell SJ Jones IM Normal prion protein has an activity like that of superoxide dismutase Biochem J 1999 344 Pt 1 1 5 10548526 10.1042/0264-6021:3440001
Mietkiewski K Domka F Malendowicz L Malendowicz J Studies on ATP hydrolysis in medium for histochemical demonstration of ATPase activity Histochemie 1970 24 343 353 4252323
Michel AD Xing M Humphrey PP Serum constituents can affect 2'-& 3'-O-(4-benzoylbenzoyl)-ATP potency at P2X(7) receptors Br J Pharmacol 2001 132 1501 1508 11264244 10.1038/sj.bjp.0703968
Brown DR Schmidt B Kretzschmar HA A neurotoxic prion protein fragment enhances proliferation of microglia but not astrocytes in culture Glia 1996 18 59 67 8891692 10.1002/(SICI)1098-1136(199609)18:1<59::AID-GLIA6>3.0.CO;2-Z
Brown DR Schmidt B Kretzschmar HA Role of microglia and host prion protein in neurotoxicity of a prion protein fragment Nature 1996 380 345 347 8598929 10.1038/380345a0
Chao CC Hu S Molitor TW Shaskan EG Peterson PK Activated microglia mediate neuronal cell injury via a nitric oxide mechanism J Immunol 1992 149 2736 2741 1383325
Dawson VL Brahmbhatt HP Mong JA Dawson TM Expression of inducible nitric oxide synthase causes delayed neurotoxicity in primary mixed neuronal-glial cortical cultures Neuropharmacology 1994 33 1425 1430 7532825 10.1016/0028-3908(94)90045-0
Demerle-Pallardy C Lonchampt MO Chabrier PE Braquet P Nitric oxide synthase induction in glial cells: effect on neuronal survival Life Sci 1993 52 1883 1890 7684804 10.1016/0024-3205(93)90009-R
Hewett SJ Csernansky CA Choi DW Selective potentiation of NMDA-induced neuronal injury following induction of astrocytic iNOS Neuron 1994 13 487 494 7520256 10.1016/0896-6273(94)90362-X
Stone R Stewart VC Hurst RD Clark JB Heales SJ Astrocyte nitric oxide causes neuronal mitochondrial damage, but antioxidant release limits neuronal cell death Ann N Y Acad Sci 1999 893 400 403 10672277
Iravani MM Kashefi K Mander P Rose S Jenner P Involvement of inducible nitric oxide synthase in inflammation-induced dopaminergic neurodegeneration Neuroscience 2002 110 49 58 11882372 10.1016/S0306-4522(01)00562-0
Ding M Zhang M Wong JL Rogers NE Ignarro LJ Voskuhl RR Antisense knockdown of inducible nitric oxide synthase inhibits induction of experimental autoimmune encephalomyelitis in SJL/J mice J Immunol 1998 160 2560 2564 9510151
David JP Ghozali F Fallet-Bianco C Wattez A Delaine S Boniface B Di Menza C Delacourte A Glial reaction in the hippocampal formation is highly correlated with aging in human brain Neurosci Lett 1997 235 53 56 9389594 10.1016/S0304-3940(97)00708-8
Ma SX Holley AT Sandra A Cassell MD Abboud FM Increased expression of nitric oxide synthase in the gracile nucleus of aged rats Neuroscience 1997 76 659 663 9135040
Vernet D Bonavera JJ Swerdloff RS Gonzalez-Cadavid NF Wang C Spontaneous expression of inducible nitric oxide synthase in the hypothalamus and other brain regions of aging rats Endocrinology 1998 139 3254 3261 9645701 10.1210/en.139.7.3254
Golde S Chandran S Brown GC Compston A Different pathways for iNOS-mediated toxicity in vitro dependent on neuronal maturation and NMDA receptor expression J Neurochem 2002 82 269 282 12124428 10.1046/j.1471-4159.2002.00973.x
Sumimoto H Kage Y Nunoi H Sasaki H Nose T Fukumaki Y Ohno M Minakami S Takeshige K Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase Proc Natl Acad Sci U S A 1994 91 5345 5349 8202490
Le Feuvre R Brough D Rothwell N Extracellular ATP and P2X7 receptors in neurodegeneration Eur J Pharmacol 2002 447 261 269 12151017 10.1016/S0014-2999(02)01848-4
Crow JP Manganese and iron porphyrins catalyze peroxynitrite decomposition and simultaneously increase nitration and oxidant yield: implications for their use as peroxynitrite scavengers in vivo Arch Biochem Biophys 1999 371 41 52 10525288 10.1006/abbi.1999.1414
Misko TP Highkin MK Veenhuizen AW Manning PT Stern MK Currie MG Salvemini D Characterization of the cytoprotective action of peroxynitrite decomposition catalysts J Biol Chem 1998 273 15646 15653 9624158 10.1074/jbc.273.25.15646
Acarin L Peluffo H Barbeito L Castellano B Gonzalez B Astroglial nitration after postnatal excitotoxic damage: correlation with nitric oxide sources, cytoskeletal, apoptotic and antioxidant proteins J Neurotrauma 2005 22 189 200 15665612 10.1089/neu.2005.22.189
Nguyen T Brunson D Crespi CL Penman BW Wishnok JS Tannenbaum SR DNA damage and mutation in human cells exposed to nitric oxide in vitro Proc Natl Acad Sci U S A 1992 89 3030 3034 1557408
Bolanos JP Heales SJ Land JM Clark JB Effect of peroxynitrite on the mitochondrial respiratory chain: differential susceptibility of neurones and astrocytes in primary culture J Neurochem 1995 64 1965 1972 7722484
Beckman JS Koppenol WH Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly Am J Physiol 1996 271 C1424 37 8944624
Leist M Fava E Montecucco C Nicotera P Peroxynitrite and nitric oxide donors induce neuronal apoptosis by eliciting autocrine excitotoxicity Eur J Neurosci 1997 9 1488 1498 9240406
Trackey JL Uliasz TF Hewett SJ SIN-1-induced cytotoxicity in mixed cortical cell culture: peroxynitrite-dependent and -independent induction of excitotoxic cell death J Neurochem 2001 79 445 455 11677273 10.1046/j.1471-4159.2001.00584.x
Prusiner SB Novel proteinaceous infectious particles cause scrapie Science 1982 216 136 144 6801762
Brown DR Kretzschmar HA Microglia and prion disease: a review Histol Histopathol 1997 12 883 892 9225170
Forloni G Angeretti N Chiesa R Monzani E Salmona M Bugiani O Tagliavini F Neurotoxicity of a prion protein fragment Nature 1993 362 543 546 8464494 10.1038/362543a0
|
16156895
|
PMC1232863
|
CC BY
|
2021-01-04 16:38:21
|
no
|
J Neuroinflammation. 2005 Sep 12; 2:20
|
utf-8
|
J Neuroinflammation
| 2,005 |
10.1186/1742-2094-2-20
|
oa_comm
|
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-391612239110.1186/1475-2875-4-39ResearchMalaria during pregnancy and foetal haematological status in Blantyre, Malawi Abrams Elizabeth T [email protected] Jesse J [email protected] Victor [email protected] Deborah D [email protected] Eyob [email protected] Valentino M [email protected] Malcolm E [email protected] Stephen J [email protected] Steven R [email protected] Department of Humanities and Social Sciences, California Institute of Technology, Pasadena, California, USA2 Department of Epidemiology, University of North Carolina, Chapel Hill, North Carolina, USA3 Department of Community Health, University College of Medicine, University of Malawi, Blantyre, Malawi4 UNC Project, Lilongwe, Malawi5 Department of Obstetrics and Gynaecology, College of Medicine, University of Malawi, Blantyre, Malawi6 School of Tropical Medicine, University of Liverpool, Liverpool, UK7 Malawi-Liverpool-Wellcome Trust Clinical Research Programme, College of Medicine, University of Malawi, Blantyre, Malawi8 Department of Medicine, University of Melbourne, Royal Melbourne Hospital, Parkville, Australia2005 25 8 2005 4 39 39 7 4 2005 25 8 2005 Copyright © 2005 Abrams et al; licensee BioMed Central Ltd.2005Abrams 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
Although maternal anaemia often stems from malaria infection during pregnancy, its effects on foetal haemoglobin levels are not straightforward. Lower-than-expected cord haemoglobin values in malarious versus non-malarious regions were noted by one review, which hypothesized they resulted from foetal immune activation to maternal malaria. This study addressed this idea by examining cord haemoglobin levels in relation to maternal malaria, anaemia, and markers of foetal immune activation.
Methods
Cord haemoglobin levels were examined in 32 malaria-infected and 58 uninfected women in Blantyre, Malawi, in relation to maternal haemoglobin levels, malaria status, and markers of foetal haematological status, hypoxia, and inflammation, including TNF-α, TGF-β, and ferritin. All women were HIV-negative.
Results
Although malaria was associated with a reduction in maternal haemoglobin (10.8 g/dL vs. 12.1 g/dL, p < 0.001), no reduction in cord haemoglobin and no significant relationship between maternal and cord haemoglobin levels were found. Cord blood markers of haematological and hypoxic statuses did not differ between malaria-infected and uninfected women. Maternal malaria was associated with decreased TGF-β and increased cord ferritin, the latter of which was positively correlated with parasitaemia (r = 0.474, p = 0.009). Increased cord ferritin was associated with significantly decreased birth weight and gestational length, although maternal and cord haemoglobin levels and malaria status had no effect on birth outcome.
Conclusion
In this population, cord haemoglobin levels were protected from the effect of maternal malaria. However, decreased TGF-β and elevated ferritin levels in cord blood suggest foetal immune activation to maternal malaria, which may help explain poor birth outcomes.
==== Body
Background
Malaria affects more than three million pregnant women per year in developing countries, where it commonly causes poor birth outcomes and maternal anaemia [1-3]. Brabin [4] suggested that foetal anaemia also resulted from maternal malaria during pregnancy. In developed countries, anaemia in newborns is rare, regardless of maternal status, and normal neonatal haemoglobin levels in these populations are higher than adult levels [e.g. [5]] In addition, there appears to be little relationship between maternal and umbilical cord haemoglobin levels, even when mothers are anaemic [e.g., [5,6]]. For example, a study of pregnant Turkish women found no significant difference in mean cord haemoglobin levels of neonates of anaemic (Hb < 10 g/dL) mothers (mean ± SEM: 16.11 ± 0.39 g/dL) compared to those of non-anaemic (Hb ≥ 12 g/dL) mothers (mean ± SEM: 16.57 ± 1.35 g/dL) [5]. In developing countries, the situation is more complex. In non-malarious areas, babies of mothers with iron-deficient anaemia have both relatively similar [e.g., [7]] and higher [e.g., [8,9]] cord haemoglobin levels than those reported for developed countries. In contrast, cord haemoglobin levels in malarious areas have been characterized by Brabin [4] as lower-than-expected, hypothesized to result from foetal immune activation to maternal malarial antigens.
In non-malarious regions, maternal anaemia may stem chiefly from iron deficiency; higher cord haemoglobin levels may thus reflect a response to foetal hypoxia caused by decreased placental oxygen transport [10]. In malarious regions, however, malaria infections during pregnancy present a second source of maternal anaemia by increasing red blood cell destruction and decreasing erythropoiesis [3,11,12]. These effects may be mediated in part by elevations in proinflammatory cytokines like TNF-α, which are associated with maternal anaemia in malaria-infected pregnancies [13,14]. Although malaria elevates placental proinflammatory cytokine production during pregnancy [14-16], it has not been reported to affect cord levels [15]. However, maternal malaria may immunologically prime the foetus, as several researchers have reported that foetal explants produce proinflammatory cytokines upon exposure to maternal malarial antigens [17,18].
To further address the hypothesis that cord haemoglobin levels in malarious areas are lower than expected due to foetal immune activation to maternal malaria, cord haemoglobin levels were examined in this study in malaria-infected and uninfected women in relation to maternal malaria status and haemoglobin levels and markers of foetal haematological, inflammatory, and hypoxic status. In specific, chronic foetal hypoxia was assessed via levels of erythropoietin (Epo) [19], a hormone responsible for red blood cell production, and corticotrophin releasing hormone (CRH) and cortisol, two hormones that have been hypothesized to mediate the effects of anaemia-related hypoxia on poor birth outcomes [20]. Tumor necrosis factor alpha (TNF-α), a pro-inflammatory cytokine, transforming growth factor beta (TGF-β), an anti-inflammatory cytokine, and C-reactive protein (CRP), an acute shock protein, were assayed as markers of inflammatory status. Haematological status was assessed via umbilical cord levels of haemoglobin (Hb) and soluble transferrin receptor (sTfR). Levels of ferritin, an iron storage protein that is an acute phase reactant in the presence of infection, were also assayed.
As markers of interest were measured both in maternal peripheral and umbilical cord blood, one critical issue is whether the cord measures truly reflect foetal production rather than placental transfer of maternally produced factors. Human data addressing this topic is limited, given the invasive and/or complex nature of the required studies. In general, studies suggest that most substances can cross the placenta, with the rate dependent on molecular weight, such that those with larger molecular weights cross slowly and those with molecular weights under 500 daltons quite rapidly [21]. Erythropoietin, a protein with a large molecular weight, is undetectable on the placental side opposite its origin [22], and it is likely that high molecular weight sequesters other large molecules like ferritin [23] on the side of the placenta on which they were produced. A placental perfusion study of inflammatory cytokine transfer suggested that TNF-α, a cytokine measured in this study, does not readily cross from the maternal to the foetal side of the placenta, although IL-6, which was not assayed in this study, appears to cross the placenta bidirectionally [24]. CRP is also unlikely to cross the placenta [25]. Cortisol may cross the placenta in its active form, but foetal exposure to maternal cortisol is normally limited by the conversion of cortisol to cortisone, its inactive metabolite, by placental 11β-hydroxysteroid dehydrogenase [26].
Methods
Subjects
Ninety pregnant women pre-labor or in the latent phase of labor attending the Labour Ward, Queen Elizabeth Central Hospital, Blantyre, Malawi, were recruited from February to October 2002 as part of a prospective cohort study investigating the impact of maternal malaria on HIV vertical transmission [27]. All women with positive peripheral malaria blood smears in the larger study were enrolled, along with the next two sequential uninfected women. Women were excluded from this study if they had HIV, preeclampsia, or multiple gestations.
Maternal socio-economic (marital status, educational level, maternal and paternal occupations, and house construction), health (antenatal clinic attendance and iron tablets usage) and anthropometric (height, weight, and mid-upper arm circumference (MUAC)) data were collected. Neonatal anthropometrics (head, abdominal, and arm circumferences, recumbent length, and weight) were evaluated within the first 24 hours after birth. Gestational age was determined by the New Ballard Score [28].
Sampling procedure
Maternal peripheral blood samples were taken at recruitment, and cord and placental samples were collected at delivery. Thick blood films were prepared from these samples for malaria microscopy. Placental blood was collected into EDTA by incising the cleaned maternal surface of the placenta and aspirating blood welling from the incision with a sterile pipette. Samples were separated within one hour, and plasma was stored at -70°C. Placental biopsies from a pericentric area (approximately 1 cm from the cord on the maternal side) were placed into 10% neutral buffered formalin to be used for placental malaria histopathology.
Malaria status
Thick blood films were air-dried and Field's stained. Plasmodium falciparum malaria parasitaemia per μl was determined by counting parasites per 200 leukocytes, assuming 6000 leukocytes per μl [29]. Placental malaria histopathology was determined by SJR as described [13]; stage of infection and degree of malaria pigment in placental monocytes and fibrin were evaluated to determine disease severity [13]. For the purposes of data analysis, malaria infection was defined as either maternal (any parasites on maternal peripheral thick blood film) or placental (any parasites on placental thick blood film or histopathology).
HIV status
All subjects were consented and received HIV pre-test and post-test counseling before the onset of active labor. HIV status was determined by Serocard rapid test for HIV-1 and 2 (Trinity Biotech, Wicklow, Ireland) and Determine HIV1/2 (Abbott Diagnostics, Abbott Park, Illinois, USA); disagreement between tests was settled by HIV-SPOT ELISA (Genelabs Diagnostics, Singapore).
Markers of haematological status
Haemoglobin (Hb) concentration were measured in peripheral and cord blood samples using a Hemocue® haemoglobinometer (HemoCue AB, Ängelholm, Sweden); some results were not available as samples clotted. Maternal peripheral and cord plasma soluble transferrin receptor (sTfR) and ferritin were assayed by enzyme-linked immunosorbent assay (ELISA) kits according to manufacturer's instructions (Ramco Laboratories, Stafford, TX; and IBL, Hamburg, Germany; respectively). Data were analysed in Microsoft Excel using a log-log standard curve. Limits of detection were as follows: 1 μg/ml (sTfR) and 0.59 ng/ml (ferritin).
Markers of inflammation
Maternal peripheral, placental and cord plasma tumor necrosis factor-α (TNF-α) and transforming growth factor-β (TGF-β) and maternal peripheral and cord plasma C reactive protein (CRP) levels were also assayed by ELISA (R&D Systems, Minneapolis, MN, USA). Color was developed by SigmaFast o-phenylenediamine dihydrochloride tablets (Sigma, St. Louis, MO, USA). Data were analysed as described above. Limits of detection were as follows: 4 pg/ml (TNF-α), 13 pg/ml (TGF-β), and 1 μg/ml (CRP).
Markers of chronic foetal hypoxia
Because scalp blood gas analysis, the most accurate means of assessing foetal hypoxia, was not feasible in the study setting, other proxies for chronic foetal hypoxia were measured: erythropoietin (Epo) [19], and cortisol and corticotrophin-releasing hormone (CRH) [20]. Maternal peripheral, placental, and cord plasma erythropoietin (Epo) levels were assayed by ELISA according to manufacturer's instructions (IBL, Hamburg, Germany). Placental and cord cortisol (CRT) and placental CRH were determined by ELISA as well (cortisol: R&D Systems, Minneapolis, MN, USA; CRH: Phoenix Pharmaceuticals, Belmont, CA, USA, respectively). Data were analysed as described above. Limits of detection were as follows: 0.4 mU/mL (EPO), 0.16 ng/ml (cortisol), and 0.19 ng/ml (CRH).
Ethical approval
This study was approved by the College of Medicine Research Committee, University of Malawi, and the Institutional Review Boards of the University of Michigan and the University of North Carolina.
Statistical analyses
Data were entered in MS-Access and analyses were performed using SPSS v.10 and Intercooled Stata 8.0. Skewed data were log or (log+1) transformed before certain analyses. Maternal characteristics in malaria-infected and uninfected women (Table 2) were compared by t-tests (continuous variables) and chi-squared tests (dichotomous variables). Mann-Whitney U tests were used to compare maternal and cord characteristics in malaria-infected and uninfected women (Tables 3 and 4). Bonferroni's method was used as a more conservative measure of significance for multiple comparisons [30]; p values reported reflect pre-adjustment values. The 5% significance level was used to determine significance.
Table 1 Malaria infections by compartment and diagnostic method
Malaria infection in any compartment Compartment Total cases Placental diagnostic method (# positive cases) Total cases
Present 42 Maternal malaria only 10
Maternal and placental malaria 19 Thick smear 1
Histopathology 5
Both 13
Placental malaria only 13 Thick smear 0
Histopathology 4
Both 9
Absent 48
Total 90
Table 2 Maternal characteristics in women with and without maternal peripheral malaria infection
Malaria-infected mean ± STD (n) total n: 29 Malaria-uninfected mean ± STD (n) total n: 61
Age 20.9 ± 3.2 (29) 22.8 ± 5.2 (61) t = 2.11, p = 0.038
Parity 1.5 ± 0.9 (29) 2.1 ± 1.5 (61) t = 2.26, p = 0.026
Number of antenatal clinic visits 4.7 ± 2.0 (29) 5.6 ± 2.3 (61) t = 1.78, p = 0.079
Maternal weight (kg) 56.4 ± 8.4 (29) 59.0 ± 7.4 (61) t = 1.51, p = 0.136
Maternal height (cm) 154.6 ± 6.0 (29) 155.3 ± 4.9 (61) t = 0.60, p = 0.552
BMI 23.6 ± 2.9 (29) 24.4 ± 2.6 (61) t = 1.44, p = 0.153
MUAC (cm) 23.7 ± 2.9 (16) 24.6 ± 2.2 (38) t = 1.33, p = 0.190
Took iron supplements 86% (25/29) 93% (55/59) χ2 = 1.16, p = 0.282
Took antimalarial tablets 83% (24/29) 95% (55/58) χ2 = 3.37, p = 0.066
Slept under mosquito net 17% (5/29) 31% (18/58) χ2 = 1.89, p = 0.169
8th grade education or less 48% (14/29) 52% (32/61) χ2 = 0.14, p = 0.711
Lived in a mud-walled house 77% (20/26) 69% (35/51) χ2 = 0.58, p = 0.446
Mother self or formally employed 21% (6/29) 20% (12/61) χ2 = 0.01, p = 0.910
Father self or formally employed 92% (23/25) 86% (44/51) χ2 = 0.53, p = 0.468
Table 3 Mann-Whitney U test of effect of maternal peripheral malaria on maternal and neonatal haematological and inflammatory status
Infected median (IQRa) n (total n: 29) Uninfected median (IQR) n (total n: 61) p
Maternal
Hb (g/dL) 10.8 (10.1 – 11.2) 29 12.1 (11.1 – 13.1)61 <0.001b
sTfR (μg/ml) 5.7 (3.8 – 9.2) 29 6.6 (4.9 – 9.7) 61 0.460
ferritin (ng/ml) 95.3 (42.3 – 180.8) 29 16.5 (6.0 – 33.1) 61 0.001b
TNF-α (pg/ml) 22.9 (7.9 – 72.1) 29 1.8 (0.0 – 24.4) 61 0.001b
TGF-β (ng/ml) 11.0 (6.0 – 15.9) 29 15.5 (9.6 – 18.3) 55 0.027
CRP (μg/ml) 85.0 (59.2 – 195.4) 29 7.6 (2.8 – 26.2) 61 <0.001b
Epo (mU/L) 60.6 (41.0 – 98.5) 29 32.7 (21.6 – 52.6) 61 0.001b
Cord
Hb (g/dL) 15.9 (14.5 – 16.9) 22 15.6 (14.5 – 17.1) 48 0.849
sTfR (μg/ml) 6.7 (4.8 – 9.9) 29 7.8 (4.9 – 9.7) 60 0.430
ferritin (ng/ml) 135.5 (66.7 – 276.3) 29 99.0 (39.4 – 158.7) 60 0.040
TNF-α (pg/ml) 1.0 (0.0 – 8.4) 28 2.7 (0.0 – 16.6) 59 0.371
TGF-β (ng/ml) 17.8 (13.5 – 24.0) 15 27.7 (20.7 – 31.9) 32 0.019
CRP (μg/ml) 1.5 (0.7 – 1.7) 28 1.0 (0.0 – 1.6) 61 0.169
Epo (mU/L) 27.1 (13.4 – 58.9) 29 25.8 (12.5 – 39.1) 61 0.641
aIQR: Inter-quartile range
b Significant after Bonferroni's adjustment for multiple comparisons.
Results
Study population
Of the 90 study participants, 29 had maternal peripheral malaria infections, 32 had placental malaria infections, and 42 had infections in one or both compartments. No cord blood malaria was identified in this study. Table 1 cross-references malaria infections by compartment and diagnostic method.
Table 2 shows the characteristics of study enrollees classified by maternal peripheral malaria. All women were HIV negative. Women with and without peripheral malaria infections were similar in terms of age and anthropometric measures. However, malaria-infected women were younger, of lower parity, and tended to have fewer antenatal clinic visits (Table 2). Socioeconomic status, as measured by educational level, house construction, and employment status, did not differ between the groups (Table 2). When women were alternatively classified by placental malaria status, there were again no significant differences in study participant characteristics (data not shown). However, when reclassified according to malaria infection in any compartment (maternal and/or placental), malaria-infected women had significantly lower BMI (23.4 ± 4.0 vs. 24.8 ± 2.4; p = 0.02) and fewer antenatal clinic visits (4.8 ± 1.8 vs. 5.7 ± 2.5; p = 0.045); no other characteristics differed between the groups (data not shown).
Data on the effects of malaria on maternal and foetal markers of haematological status, inflammation, and chronic hypoxia are presented in two tables: Table 3 presents the effects of maternal peripheral malaria on the markers, and Table 4 presents the effects of placental malaria.
Table 4 Mann-Whitney U test of effect of placental malaria on markers of neonatal haematological, inflammatory, and hypoxic status
Infected median (IQRa) n (total n: 32) Uninfected median (IQR) n (total n: 58)
Cord
Hb (g/dL) 15.4 (13.7 – 16.8) 25 16.0 (15.0 – 17.1) 45 p = 0.128
sTfR (μg/ml) 6.7 (5.1 – 9.9) 32 7.7 (4.3 – 9.6) 58 p = 0.765
ferritin (ng/ml) 129.0 (61.9 – 248.6) 32 101.0 (41.2 – 210.8) 57 p = 0.184
TNF-α (pg/ml) 1.7 (0.0 – 5.6) 30 2.7 (0.0 – 15.9) 57 p = 0.552
TGF-β (ng/ml) 19.5 (15.6 – 30.3) 16 26.8 (17.9 – 31.8) 31 p = 0.080
CRP (μg/ml) 1.5 (0.7 – 1.6) 31 1.0 (0.0–1.6) 58 p = 0.289
Epo (mU/L) 25.4 (13.4 – 58.1) 32 27.5 (12.5 – 45.1) 58 p = 0.940
Cortisol (μg/ml) 421.0 (283.5 – 538.5) 32 511.0 (324.0 – 677.5) 53 p = 0.425
Placenta
Cortisol (μg/ml) 191.0 (155.5 – 251.0) 29 172.0 (132.0 – 248.5) 53 p = 0.357
CRH (pg/ml) 90.0 (42.5 – 126.4) 24 72.3 (41.5 – 116.3) 43 p = 0.421
aIQR: Inter-quartile range
The effect of malaria on haematological status differed markedly between women and neonates. Compared to their uninfected counterparts, women characterized by any of the three definitions of malaria infection tended to have significantly lower maternal Hb concentrations (maternal: Table 3; placental: p = 0.017; any: p < 0.001). Unlike maternal haematological status, neonatal characteristics were protected from the effects of maternal and placental malaria. Mean cord Hb levels did not differ between malaria-infected and uninfected women (Tables 3 and 4). In addition, malarial disease severity, including peripheral and placental parasite density, stage of disease, and deposition of fibrin and malaria pigment in placental monocytes, had no effect on neonatal Hb (data not shown). Neither cord nor maternal sTfR levels were affected by malaria infection in this study (Tables 3 and 4).
Markers of inflammation were also altered in malaria-infected women, who had significantly elevated CRP and TNF-α (maternal: Table 3; placental: CRP (p < 0.001), TNF-α (p = 0.022); any: CRP (p < 0.001), TNF-α (p = 0.004)) and lower TGF-β, although this result was not significant after Bonferroni's adjustment (maternal: Table 3; placental: NS; any: p = 0.010). There was no effect of maternal malaria on foetal TNF-α or CRP (Table 3). TGF-β was significantly decreased in the cord blood of neonates with peripheral malaria-infected mothers, although this result was not significant after Bonferroni's adjustment (Table 3). Among babies of malaria-infected women, neither increased maternal or placental parasite density nor markers of disease severity were associated with increases in measures of neonatal inflammation (data not shown). However, elevated maternal CRP in women with peripheral malaria infections was associated with significantly increased cord CRP (r = 0.48, p = 0.010) and decreased cord TGF-β (r = -0.60, p = 0.019), suggesting an association between maternal and foetal immune activation.
Malaria-infected women had significantly higher ferritin levels than uninfected women (maternal: Table 3; placental: p < 0.001; any: p < 0.001). Cord ferritin was significantly increased in the cord blood of neonates with peripheral malaria-infected mothers, although this result was not significant after Bonferroni's adjustment (Table 2). In addition, cord ferritin was correlated with increasing maternal parasite load in malaria-infected women (r = 0.474, p = 0.009). The notion that ferritin levels in this group reflected malaria-related inflammation is further supported by the inverse correlation between cord ferritin and cord TGF-β (r = -.593, p = 0.020). Additionally, primigravids with placental malaria infections, who tend to experience its inflammatory effects more acutely than multigravids, had significantly higher cord ferritin levels than did multigravid malaria-infected women (p = 0.021). However, despite the indications of immune activation in neonates of malaria-infected women, none of the cord immune markers were correlated with cord Hb or sTfR (data not shown).
Although women characterized by any of the three definitions of malaria infection tended to have significantly higher Epo levels than uninfected women (maternal: Table 3; placental: NS; any: p = 0.004), none of the proxies for foetal hypoxia examined in the study (cord Epo, placental and cord cortisol, and placental CRH) were increased in the neonates of malaria-infected women (Tables 3 and 4). Among malaria-infected women, increased maternal and placental parasitaemia and markers of disease severity were also not associated with increased levels of these factors (data not shown).
Neonatal haematological, immune and hypoxic markers were then examined in relation to poor birth outcomes that have been associated with maternal malaria, particularly low birthweight and preterm delivery. In this study, neither the presence nor the severity of maternal or placental malaria was associated with decreased birthweight or gestational age, perhaps due to the limited sample size (data not shown). Unsurprisingly, therefore, neither decreased maternal nor cord Hb was significantly associated with poorer birth outcomes in malaria-infected women (data not shown), although cord sTfR levels, which were not different between neonates of malaria-infected and uninfected women (Tables 3 and 4), were inversely associated with birthweight in women with maternal, but not placental, malaria (r = -.452, p = 0.014). In neonates of women with both maternal and placental malaria, elevated cord ferritin was correlated with lower birthweights (maternal: r = -0.427, p = 0.021; placental: r = -0.374, p = 0.035), and in those of women with maternal malaria, with shorter gestations as well (r = -0.432, p = 0.024). This association was not due to inflammation alone, since there was no association with cord CRP or TGF-β, nor to foetal haematopoiesis alone, since there was no association with cord Epo (data not shown). However, the data suggest that a combination of the haematological and inflammatory processes might be associated with poor birth outcomes commonly found following malaria infection during pregnancy.
Discussion
In this study, the haematological status of neonates of mothers with malaria parasitaemia at delivery was not significantly impacted. Despite the effect of malaria on maternal Hb, which was expectedly lower, neonatal Hb levels were not decreased in the cord blood of malaria-infected compared to uninfected women. In addition, cord Hb values were not correlated to maternal Hb concentration in either group or severity of malaria, as described in the Results section. Although the sample size was relatively small, these results resemble more closely the results from Western studies, which tend to find unaffected cord values, than the studies of malaria- and non-malaria-exposed mother-neonate pairs in developing countries, as reviewed by Brabin [4]. It is unclear why, but the high cord haemoglobin levels observed in this study may relate to the high rates of antimalarial usage and iron supplementation in Malawian women (Table 2), which might provide a protective effect. The maintenance of cord Hb levels despite the presence of maternal anaemia and malaria suggests that the foetus has developed mechanisms to preferentially obtain sufficient iron and produce adequate amounts of red cells. Iron is transported unidirectionally from mother to foetus across a concentration gradient [31], and thus stores should be preferentially preserved in the foetus.
Ferritin is an iron storage protein that is an acute phase reactant in the presence of infection. While cord ferritin decreases in foetuses of anaemic mothers without malaria, it appears to increase in foetuses of mothers with increasing parasitaemia. Increased ferritin in the presence of inflammation has been suggested to reflect an evolutionary strategy by which the body removes bioavailable iron from the presence of parasites [32]. In foetuses of mothers with malaria, it may have an additional function: to help sequester iron obtained from the mother. Interestingly, in this study, elevated cord ferritin levels best predicted poor birth outcomes, including shorter gestation and lower birth weight. It is unclear why this association was found, particularly since no association was found in this study between birth outcomes and malaria infection or maternal anaemia, two well-established risk factors. This suggests that increased cord ferritin might be a result of the same foetal stresses that cause poor birth outcomes.
Many researchers have assumed that foetal hypoxia, rather than inflammation, mediates the effect of malaria and maternal anaemia on birth outcomes, particularly low birth weight. Despite the induction of foetal hypoxia by maternal anaemia [20,33] and the decrease in umbilical blood flow in pregnant women with malaria [34], it has not been rigorously demonstrated that chronic low-grade hypoxia causes poor birth outcomes, and it is unclear how severe anaemia must be to induce foetal hypoxia. Foetal hypoxia is difficult to quantify in nonwestern settings. In developed countries, blood gases are collected as soon as the foetal scalp is discernable exiting the birth canal. This technique is not feasible in hospitals in developing countries, but several biochemical measures in cord blood may offer proxies for foetal hypoxia. In particular, Epo, the hormone primarily responsible for regulation of erythropoiesis, is stimulated by both anaemia and hypoxia [35] and is a good marker of chronic foetal hypoxia [19,36] because Epo does not cross the placenta. Several studies have reported both adequate [37-39] and inadequate [40,41] Epo production for the level of anaemia in malaria-infected children and adults, and this question requires further investigation. In this study, maternal Epo was elevated in malaria-infected women (Table 3), but cord Epo levels were not affected by either maternal or placental malaria infections (Tables 3 and 4).
Allen [20] hypothesized that foetal stress hormones (cortisol and CRH) might mediate the link between maternal anaemia and poor birth outcomes, via the effect of anaemia-induced foetal hypoxia on these hormones. In this study, cord cortisol and placental cortisol and CRH were neither elevated in association with maternal malaria infection nor associated with poor birth outcomes. This may reflect the inadequacy of these factors as markers for chronic foetal hypoxia, either in general or in cord blood, or may simply be a result of the small sample size. On the other hand, hypoxia in the foetus may be so fleeting, as the foetus downregulates its growth in relation to available oxygen levels, that it does not trigger lasting detectable markers. This response is seen in sheep, where pregnant females with experimentally-induced anaemia evince a compensatory decrease in foetal growth but no actual evidence of hypoxia [42].
Conclusion
In summary, cord haemoglobin levels were unchanged by maternal malaria and anaemia during pregnancy in this population. However, maternal parasitaemia induced a foetal inflammatory response, specifically lower TGF-β and higher ferritin levels. Elevated cord ferritin in turn was associated with lower birthweights and shorter gestations.
Authors' contributions
ETA was responsible for conception and design of the study; IRB approval; conducting assays; analysis and interpretation of data; drafting the paper; and revising it for publication. JK and DDK performed many of the assays and analyses of these assays. VM designed the larger study and conducted much of the data analysis of that study that formed the foundation of this one. ET and VLM supervised patient enrollment and acquisition of data and aided in its submission to the IRB. MEM and SRM were central to the conception and design of the study, the analysis and interpretation of data, and the revision of the paper for publication. SRM was involved in the conception and design of the study; IRB approval; analysis and interpretation of data; drafting the paper; and revising the paper for publication.
Acknowledgements
Funding for this study was provided by The Wenner-Gren Foundation for Anthropological Research (ETA), the Anthropology Department of the University of Michigan (ETA), the Rackham School of Graduate Studies at the University of Michigan (ETA), the Wellcome Trust Career Development Fellowship (SJR; ref 046012), and the NIH (SRM; AI 49084).
==== Refs
Brabin BJ An analysis of malaria in pregnancy in Africa Bull World Health Organ 1983 61 1005 1016 6370484
Menendez C Ordi J Ismail MR Ventura PJ Aponte JJ Kahigwa E Font F Alonso PL The impact of placental malaria on gestational age and birth weight J Infect Dis 2000 181 1740 1745 10823776 10.1086/315449
Menendez C Fleming AF Alonso PL Malaria-related anaemia Parasitol Today 2000 16 469 476 11063857 10.1016/S0169-4758(00)01774-9
Brabin B Foetal anaemia in malarious areas: its causes and significance Ann Trop Paediatr 1992 12 303 310 1280047
Erdem A Erdem M Arslan M Yazici G Eskandari R Himmetoglu O The effect of maternal anaemia and iron deficiency on fetal erythropoiesis: comparison between serum erythropoietin, haemoglobin and ferritin levels in mothers and newborns J Matern Fetal Neonatal Med 2002 11 329 332 12389675
Hokama T Takenaka S Hirayama K Yara A Yoshida K Itokazu K Kinjho R Yabu E Iron status of newborns born to iron deficient anaemic mothers J Trop Pediatr 1996 42 75 77 8984218 10.1093/tropej/42.2.75
Agarwal KN Krishna M Singla PN Jain A Effects of hematinic and protein supplementation on the feto-placental unit in maternal anaemia Indian Pediatr 1984 21 765 771 6526483
Tracer DP Reproductive and socio-economic correlates of maternal haemoglobin levels in a rural area of Papua New Guinea Trop Med Int Health 1997 2 513 518 9236817 10.1046/j.1365-3156.1997.d01-327.x
Bhargava M Iyer PU Kumar R Ramji S Kapani V Bhargava SK Relationship of maternal serum ferritin with foetal serum ferritin, birth weight and gestation J Trop Pediatr 1991 37 149 152 1960769
Ballew C Haas JD Haematologic evidence of foetal hypoxia among newborn infants at high altitude in Bolivia Am J Obstet Gynecol 1986 155 166 169 2425621
Abdalla SH Haematopoiesis in human malaria Blood Cells 1990 16 401 416 discussion 417-409 2257320
Huddle JM Gibson RS Cullinan TR The impact of malarial infection and diet on the anaemia status of rural pregnant Malawian women Eur J Clin Nutr 1999 53 792 801 10556986 10.1038/sj.ejcn.1600851
Rogerson SJ Pollina E Getachew A Tadesse E Lema VM Molyneux ME Placental monocyte infiltrates in response to Plasmodium falciparum malaria infection and their association with adverse pregnancy outcomes Am J Trop Med Hyg 2003 68 115 119 12556159
Fried M Muga RO Misore AO Duffy PE Malaria elicits type 1 cytokines in the human placenta: IFN-gamma and TNF-alpha associated with pregnancy outcomes J Immunol 1998 160 2523 2530 9498798
Rogerson SJ Brown HC Pollina E Abrams ET Tadesse E Lema VM Molyneux ME Placental tumor necrosis factor alpha but not gamma interferon is associated with placental malaria and low birth weight in Malawian women Infect Immun 2003 71 267 270 12496175 10.1128/IAI.71.1.267-270.2003
Moormann AM Sullivan AD Rochford RA Chensue SW Bock PJ Nyirenda T Meshnick SR Malaria and pregnancy: placental cytokine expression and its relationship to intrauterine growth retardation J Infect Dis 1999 180 1987 1993 10558956 10.1086/315135
Fievet N Ringwald P Bickii J Dubois B Maubert B Le Hesran JY Cot M Deloron P Malaria cellular immune responses in neonates from Cameroon Parasite Immunol 1996 18 483 490 9226685 10.1046/j.1365-3024.1996.d01-19.x
King CL Malhotra I Wamachi A Kioko J Mungai P Wahab SA Koech D Zimmerman P Ouma J Kazura JW Acquired immune responses to Plasmodium falciparum merozoite surface protein-1 in the human foetus J Immunol 2002 168 356 364 11751981
Buescher U Hertwig K Wolf C Dudenhausen JW Erythropoietin in amniotic fluid as a marker of chronic foetal hypoxia Int J Gynaecol Obstet 1998 60 257 263 9544710 10.1016/S0020-7292(98)00011-3
Allen LH Biological mechanisms that might underlie iron's effects on foetal growth and preterm birth J Nutr 2001 131 581S 589S 11160591
Pacifici GM Nottoli R Placental transfer of drugs administered to the mother Clin Pharmacokinet 1995 28 235 269 7758253
Zanjani ED Pixley JS Slotnick N MacKintosh FR Ekhterae D Clemons G Erythropoietin does not cross the placenta into the foetus Pathobiology 1993 61 211 215 8216843
Starreveld JS Kroos MJ Van Suijlen JD Verrijt CE Van Eijk HG Van Dijk JP Ferritin in cultured human cytotrophoblasts: synthesis and subunit distribution Placenta 1995 16 383 395 7567801 10.1016/0143-4004(95)90096-9
Zaretsky MV Alexander JM Byrd W Bawdon RE Transfer of inflammatory cytokines across the placenta Obstet Gynecol 2004 103 546 550 14990420
Gutteberg TJ Askvik K Jorgensen T Serum lactoferrin and C-reactive protein in mother and newborn after preterm rupture of membranes Acta Obstet Gynecol Scand 1986 65 203 205 3488635
Yang K Placental 11 beta-hydroxysteroid dehydrogenase: barrier to maternal glucocorticoids Rev Reprod 1997 2 129 132 9414475 10.1530/ror.0.0020129
Mwapasa V Rogerson SJ Molyneux ME Abrams ET Kamwendo DD Lema VM Tadesse E Chaluluka E Wilson PE Meshnick SR The effect of Plasmodium falciparum malaria on peripheral and placental HIV-1 RNA concentrations in pregnant Malawian women Aids 2004 18 1051 1059 15096809 10.1097/00002030-200404300-00014
Ballard JL Khoury KC Wedig K Wang L Eilers-Walsman BL Lipp R New Ballard Score, expanded to include extremely premature infants J Pediatr 1991 119 417 423 1880657
Rogerson SJ Chaluluka E Kanjala M Mkundika P Mhango C Molyneux ME Intermittent sulfadoxine-pyrimethamine in pregnancy: effectiveness against malaria morbidity in Blantyre, Malawi, in 1997–99 Trans R Soc Trop Med Hyg 2000 94 549 553 11132387 10.1016/S0035-9203(00)90083-X
Sokal RR Rohlf FJ Biometry 1995 New York: W. H. Freeman and Company
Fletcher J Suter PE The transport of iron by the human placenta Clin Sci 1969 36 209 220 5772100
Weinberg ED Nutritional immunity. Host's attempt to withold iron from microbial invaders Jama 1975 231 39 41 1243565 10.1001/jama.231.1.39
Rondo PH Tomkins AM Maternal iron status and intrauterine growth retardation Trans R Soc Trop Med Hyg 1999 93 423 426 10674095 10.1016/S0035-9203(99)90144-X
Arbeille P Carles G Bousquet F Body G Lansac J Foetal cerebral and umbilical artery blood flow changes during pregnancy complicated by malaria J Ultrasound Med 1998 17 223 229 9544605
Vora M Gruslin A Erythropoietin in obstetrics Obstet Gynecol Surv 1998 53 500 508 9702790 10.1097/00006254-199808000-00023
Ostlund E Lindholm H Hemsen A Fried G Foetal erythropoietin and endothelin-1: relation to hypoxia and intrauterine growth retardation Acta Obstet Gynecol Scand 2000 79 276 282 10746842 10.1034/j.1600-0412.2000.079004276.x
Burchard GD Radloff P Philipps J Nkeyi M Knobloch J Kremsner PG Increased erythropoietin production in children with severe malarial anaemia Am J Trop Med Hyg 1995 53 547 551 7485716
Nussenblatt V Mukasa G Metzger A Ndeezi G Garrett E Semba RD Anaemia and interleukin-10, tumor necrosis factor alpha, and erythropoietin levels among children with acute, uncomplicated Plasmodium falciparum malaria Clin Diagn Lab Immunol 2001 8 1164 1170 11687458 10.1128/CDLI.8.6.1164-1170.2001
Verhoef H West CE Kraaijenhagen R Nzyuko SM King R Mbandi MM van Laatum S Hogervorst R Schep C Kok FJ Malarial anaemia leads to adequately increased erythropoiesis in asymptomatic Kenyan children Blood 2002 100 3489 3494 12393621 10.1182/blood-2001-12-0228
el Hassan AM Saeed AM Fandrey J Jelkmann W Decreased erythropoietin response in Plasmodium falciparum malaria-associated anaemia Eur J Haematol 1997 59 299 304 9414641
Vedovato M De Paoli Vitali E Dapporto M Salvatorelli G Defective erythropoietin production in the anaemia of malaria Nephrol Dial Transplant 1999 14 1043 1044 10328513 10.1093/ndt/14.4.1043
Mostello D Chalk C Khoury J Mack CE Siddiqi TA Clark KE Chronic anaemia in pregnant ewes: maternal and foetal effects Am J Physiol 1991 261 R1075 1083 1951756
|
16122391
|
PMC1232864
|
CC BY
|
2021-01-04 16:24:13
|
no
|
Malar J. 2005 Aug 25; 4:39
|
utf-8
|
Malar J
| 2,005 |
10.1186/1475-2875-4-39
|
oa_comm
|
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-391612239110.1186/1475-2875-4-39ResearchMalaria during pregnancy and foetal haematological status in Blantyre, Malawi Abrams Elizabeth T [email protected] Jesse J [email protected] Victor [email protected] Deborah D [email protected] Eyob [email protected] Valentino M [email protected] Malcolm E [email protected] Stephen J [email protected] Steven R [email protected] Department of Humanities and Social Sciences, California Institute of Technology, Pasadena, California, USA2 Department of Epidemiology, University of North Carolina, Chapel Hill, North Carolina, USA3 Department of Community Health, University College of Medicine, University of Malawi, Blantyre, Malawi4 UNC Project, Lilongwe, Malawi5 Department of Obstetrics and Gynaecology, College of Medicine, University of Malawi, Blantyre, Malawi6 School of Tropical Medicine, University of Liverpool, Liverpool, UK7 Malawi-Liverpool-Wellcome Trust Clinical Research Programme, College of Medicine, University of Malawi, Blantyre, Malawi8 Department of Medicine, University of Melbourne, Royal Melbourne Hospital, Parkville, Australia2005 25 8 2005 4 39 39 7 4 2005 25 8 2005 Copyright © 2005 Abrams et al; licensee BioMed Central Ltd.2005Abrams 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
Although maternal anaemia often stems from malaria infection during pregnancy, its effects on foetal haemoglobin levels are not straightforward. Lower-than-expected cord haemoglobin values in malarious versus non-malarious regions were noted by one review, which hypothesized they resulted from foetal immune activation to maternal malaria. This study addressed this idea by examining cord haemoglobin levels in relation to maternal malaria, anaemia, and markers of foetal immune activation.
Methods
Cord haemoglobin levels were examined in 32 malaria-infected and 58 uninfected women in Blantyre, Malawi, in relation to maternal haemoglobin levels, malaria status, and markers of foetal haematological status, hypoxia, and inflammation, including TNF-α, TGF-β, and ferritin. All women were HIV-negative.
Results
Although malaria was associated with a reduction in maternal haemoglobin (10.8 g/dL vs. 12.1 g/dL, p < 0.001), no reduction in cord haemoglobin and no significant relationship between maternal and cord haemoglobin levels were found. Cord blood markers of haematological and hypoxic statuses did not differ between malaria-infected and uninfected women. Maternal malaria was associated with decreased TGF-β and increased cord ferritin, the latter of which was positively correlated with parasitaemia (r = 0.474, p = 0.009). Increased cord ferritin was associated with significantly decreased birth weight and gestational length, although maternal and cord haemoglobin levels and malaria status had no effect on birth outcome.
Conclusion
In this population, cord haemoglobin levels were protected from the effect of maternal malaria. However, decreased TGF-β and elevated ferritin levels in cord blood suggest foetal immune activation to maternal malaria, which may help explain poor birth outcomes.
==== Body
Background
Malaria affects more than three million pregnant women per year in developing countries, where it commonly causes poor birth outcomes and maternal anaemia [1-3]. Brabin [4] suggested that foetal anaemia also resulted from maternal malaria during pregnancy. In developed countries, anaemia in newborns is rare, regardless of maternal status, and normal neonatal haemoglobin levels in these populations are higher than adult levels [e.g. [5]] In addition, there appears to be little relationship between maternal and umbilical cord haemoglobin levels, even when mothers are anaemic [e.g., [5,6]]. For example, a study of pregnant Turkish women found no significant difference in mean cord haemoglobin levels of neonates of anaemic (Hb < 10 g/dL) mothers (mean ± SEM: 16.11 ± 0.39 g/dL) compared to those of non-anaemic (Hb ≥ 12 g/dL) mothers (mean ± SEM: 16.57 ± 1.35 g/dL) [5]. In developing countries, the situation is more complex. In non-malarious areas, babies of mothers with iron-deficient anaemia have both relatively similar [e.g., [7]] and higher [e.g., [8,9]] cord haemoglobin levels than those reported for developed countries. In contrast, cord haemoglobin levels in malarious areas have been characterized by Brabin [4] as lower-than-expected, hypothesized to result from foetal immune activation to maternal malarial antigens.
In non-malarious regions, maternal anaemia may stem chiefly from iron deficiency; higher cord haemoglobin levels may thus reflect a response to foetal hypoxia caused by decreased placental oxygen transport [10]. In malarious regions, however, malaria infections during pregnancy present a second source of maternal anaemia by increasing red blood cell destruction and decreasing erythropoiesis [3,11,12]. These effects may be mediated in part by elevations in proinflammatory cytokines like TNF-α, which are associated with maternal anaemia in malaria-infected pregnancies [13,14]. Although malaria elevates placental proinflammatory cytokine production during pregnancy [14-16], it has not been reported to affect cord levels [15]. However, maternal malaria may immunologically prime the foetus, as several researchers have reported that foetal explants produce proinflammatory cytokines upon exposure to maternal malarial antigens [17,18].
To further address the hypothesis that cord haemoglobin levels in malarious areas are lower than expected due to foetal immune activation to maternal malaria, cord haemoglobin levels were examined in this study in malaria-infected and uninfected women in relation to maternal malaria status and haemoglobin levels and markers of foetal haematological, inflammatory, and hypoxic status. In specific, chronic foetal hypoxia was assessed via levels of erythropoietin (Epo) [19], a hormone responsible for red blood cell production, and corticotrophin releasing hormone (CRH) and cortisol, two hormones that have been hypothesized to mediate the effects of anaemia-related hypoxia on poor birth outcomes [20]. Tumor necrosis factor alpha (TNF-α), a pro-inflammatory cytokine, transforming growth factor beta (TGF-β), an anti-inflammatory cytokine, and C-reactive protein (CRP), an acute shock protein, were assayed as markers of inflammatory status. Haematological status was assessed via umbilical cord levels of haemoglobin (Hb) and soluble transferrin receptor (sTfR). Levels of ferritin, an iron storage protein that is an acute phase reactant in the presence of infection, were also assayed.
As markers of interest were measured both in maternal peripheral and umbilical cord blood, one critical issue is whether the cord measures truly reflect foetal production rather than placental transfer of maternally produced factors. Human data addressing this topic is limited, given the invasive and/or complex nature of the required studies. In general, studies suggest that most substances can cross the placenta, with the rate dependent on molecular weight, such that those with larger molecular weights cross slowly and those with molecular weights under 500 daltons quite rapidly [21]. Erythropoietin, a protein with a large molecular weight, is undetectable on the placental side opposite its origin [22], and it is likely that high molecular weight sequesters other large molecules like ferritin [23] on the side of the placenta on which they were produced. A placental perfusion study of inflammatory cytokine transfer suggested that TNF-α, a cytokine measured in this study, does not readily cross from the maternal to the foetal side of the placenta, although IL-6, which was not assayed in this study, appears to cross the placenta bidirectionally [24]. CRP is also unlikely to cross the placenta [25]. Cortisol may cross the placenta in its active form, but foetal exposure to maternal cortisol is normally limited by the conversion of cortisol to cortisone, its inactive metabolite, by placental 11β-hydroxysteroid dehydrogenase [26].
Methods
Subjects
Ninety pregnant women pre-labor or in the latent phase of labor attending the Labour Ward, Queen Elizabeth Central Hospital, Blantyre, Malawi, were recruited from February to October 2002 as part of a prospective cohort study investigating the impact of maternal malaria on HIV vertical transmission [27]. All women with positive peripheral malaria blood smears in the larger study were enrolled, along with the next two sequential uninfected women. Women were excluded from this study if they had HIV, preeclampsia, or multiple gestations.
Maternal socio-economic (marital status, educational level, maternal and paternal occupations, and house construction), health (antenatal clinic attendance and iron tablets usage) and anthropometric (height, weight, and mid-upper arm circumference (MUAC)) data were collected. Neonatal anthropometrics (head, abdominal, and arm circumferences, recumbent length, and weight) were evaluated within the first 24 hours after birth. Gestational age was determined by the New Ballard Score [28].
Sampling procedure
Maternal peripheral blood samples were taken at recruitment, and cord and placental samples were collected at delivery. Thick blood films were prepared from these samples for malaria microscopy. Placental blood was collected into EDTA by incising the cleaned maternal surface of the placenta and aspirating blood welling from the incision with a sterile pipette. Samples were separated within one hour, and plasma was stored at -70°C. Placental biopsies from a pericentric area (approximately 1 cm from the cord on the maternal side) were placed into 10% neutral buffered formalin to be used for placental malaria histopathology.
Malaria status
Thick blood films were air-dried and Field's stained. Plasmodium falciparum malaria parasitaemia per μl was determined by counting parasites per 200 leukocytes, assuming 6000 leukocytes per μl [29]. Placental malaria histopathology was determined by SJR as described [13]; stage of infection and degree of malaria pigment in placental monocytes and fibrin were evaluated to determine disease severity [13]. For the purposes of data analysis, malaria infection was defined as either maternal (any parasites on maternal peripheral thick blood film) or placental (any parasites on placental thick blood film or histopathology).
HIV status
All subjects were consented and received HIV pre-test and post-test counseling before the onset of active labor. HIV status was determined by Serocard rapid test for HIV-1 and 2 (Trinity Biotech, Wicklow, Ireland) and Determine HIV1/2 (Abbott Diagnostics, Abbott Park, Illinois, USA); disagreement between tests was settled by HIV-SPOT ELISA (Genelabs Diagnostics, Singapore).
Markers of haematological status
Haemoglobin (Hb) concentration were measured in peripheral and cord blood samples using a Hemocue® haemoglobinometer (HemoCue AB, Ängelholm, Sweden); some results were not available as samples clotted. Maternal peripheral and cord plasma soluble transferrin receptor (sTfR) and ferritin were assayed by enzyme-linked immunosorbent assay (ELISA) kits according to manufacturer's instructions (Ramco Laboratories, Stafford, TX; and IBL, Hamburg, Germany; respectively). Data were analysed in Microsoft Excel using a log-log standard curve. Limits of detection were as follows: 1 μg/ml (sTfR) and 0.59 ng/ml (ferritin).
Markers of inflammation
Maternal peripheral, placental and cord plasma tumor necrosis factor-α (TNF-α) and transforming growth factor-β (TGF-β) and maternal peripheral and cord plasma C reactive protein (CRP) levels were also assayed by ELISA (R&D Systems, Minneapolis, MN, USA). Color was developed by SigmaFast o-phenylenediamine dihydrochloride tablets (Sigma, St. Louis, MO, USA). Data were analysed as described above. Limits of detection were as follows: 4 pg/ml (TNF-α), 13 pg/ml (TGF-β), and 1 μg/ml (CRP).
Markers of chronic foetal hypoxia
Because scalp blood gas analysis, the most accurate means of assessing foetal hypoxia, was not feasible in the study setting, other proxies for chronic foetal hypoxia were measured: erythropoietin (Epo) [19], and cortisol and corticotrophin-releasing hormone (CRH) [20]. Maternal peripheral, placental, and cord plasma erythropoietin (Epo) levels were assayed by ELISA according to manufacturer's instructions (IBL, Hamburg, Germany). Placental and cord cortisol (CRT) and placental CRH were determined by ELISA as well (cortisol: R&D Systems, Minneapolis, MN, USA; CRH: Phoenix Pharmaceuticals, Belmont, CA, USA, respectively). Data were analysed as described above. Limits of detection were as follows: 0.4 mU/mL (EPO), 0.16 ng/ml (cortisol), and 0.19 ng/ml (CRH).
Ethical approval
This study was approved by the College of Medicine Research Committee, University of Malawi, and the Institutional Review Boards of the University of Michigan and the University of North Carolina.
Statistical analyses
Data were entered in MS-Access and analyses were performed using SPSS v.10 and Intercooled Stata 8.0. Skewed data were log or (log+1) transformed before certain analyses. Maternal characteristics in malaria-infected and uninfected women (Table 2) were compared by t-tests (continuous variables) and chi-squared tests (dichotomous variables). Mann-Whitney U tests were used to compare maternal and cord characteristics in malaria-infected and uninfected women (Tables 3 and 4). Bonferroni's method was used as a more conservative measure of significance for multiple comparisons [30]; p values reported reflect pre-adjustment values. The 5% significance level was used to determine significance.
Table 1 Malaria infections by compartment and diagnostic method
Malaria infection in any compartment Compartment Total cases Placental diagnostic method (# positive cases) Total cases
Present 42 Maternal malaria only 10
Maternal and placental malaria 19 Thick smear 1
Histopathology 5
Both 13
Placental malaria only 13 Thick smear 0
Histopathology 4
Both 9
Absent 48
Total 90
Table 2 Maternal characteristics in women with and without maternal peripheral malaria infection
Malaria-infected mean ± STD (n) total n: 29 Malaria-uninfected mean ± STD (n) total n: 61
Age 20.9 ± 3.2 (29) 22.8 ± 5.2 (61) t = 2.11, p = 0.038
Parity 1.5 ± 0.9 (29) 2.1 ± 1.5 (61) t = 2.26, p = 0.026
Number of antenatal clinic visits 4.7 ± 2.0 (29) 5.6 ± 2.3 (61) t = 1.78, p = 0.079
Maternal weight (kg) 56.4 ± 8.4 (29) 59.0 ± 7.4 (61) t = 1.51, p = 0.136
Maternal height (cm) 154.6 ± 6.0 (29) 155.3 ± 4.9 (61) t = 0.60, p = 0.552
BMI 23.6 ± 2.9 (29) 24.4 ± 2.6 (61) t = 1.44, p = 0.153
MUAC (cm) 23.7 ± 2.9 (16) 24.6 ± 2.2 (38) t = 1.33, p = 0.190
Took iron supplements 86% (25/29) 93% (55/59) χ2 = 1.16, p = 0.282
Took antimalarial tablets 83% (24/29) 95% (55/58) χ2 = 3.37, p = 0.066
Slept under mosquito net 17% (5/29) 31% (18/58) χ2 = 1.89, p = 0.169
8th grade education or less 48% (14/29) 52% (32/61) χ2 = 0.14, p = 0.711
Lived in a mud-walled house 77% (20/26) 69% (35/51) χ2 = 0.58, p = 0.446
Mother self or formally employed 21% (6/29) 20% (12/61) χ2 = 0.01, p = 0.910
Father self or formally employed 92% (23/25) 86% (44/51) χ2 = 0.53, p = 0.468
Table 3 Mann-Whitney U test of effect of maternal peripheral malaria on maternal and neonatal haematological and inflammatory status
Infected median (IQRa) n (total n: 29) Uninfected median (IQR) n (total n: 61) p
Maternal
Hb (g/dL) 10.8 (10.1 – 11.2) 29 12.1 (11.1 – 13.1)61 <0.001b
sTfR (μg/ml) 5.7 (3.8 – 9.2) 29 6.6 (4.9 – 9.7) 61 0.460
ferritin (ng/ml) 95.3 (42.3 – 180.8) 29 16.5 (6.0 – 33.1) 61 0.001b
TNF-α (pg/ml) 22.9 (7.9 – 72.1) 29 1.8 (0.0 – 24.4) 61 0.001b
TGF-β (ng/ml) 11.0 (6.0 – 15.9) 29 15.5 (9.6 – 18.3) 55 0.027
CRP (μg/ml) 85.0 (59.2 – 195.4) 29 7.6 (2.8 – 26.2) 61 <0.001b
Epo (mU/L) 60.6 (41.0 – 98.5) 29 32.7 (21.6 – 52.6) 61 0.001b
Cord
Hb (g/dL) 15.9 (14.5 – 16.9) 22 15.6 (14.5 – 17.1) 48 0.849
sTfR (μg/ml) 6.7 (4.8 – 9.9) 29 7.8 (4.9 – 9.7) 60 0.430
ferritin (ng/ml) 135.5 (66.7 – 276.3) 29 99.0 (39.4 – 158.7) 60 0.040
TNF-α (pg/ml) 1.0 (0.0 – 8.4) 28 2.7 (0.0 – 16.6) 59 0.371
TGF-β (ng/ml) 17.8 (13.5 – 24.0) 15 27.7 (20.7 – 31.9) 32 0.019
CRP (μg/ml) 1.5 (0.7 – 1.7) 28 1.0 (0.0 – 1.6) 61 0.169
Epo (mU/L) 27.1 (13.4 – 58.9) 29 25.8 (12.5 – 39.1) 61 0.641
aIQR: Inter-quartile range
b Significant after Bonferroni's adjustment for multiple comparisons.
Results
Study population
Of the 90 study participants, 29 had maternal peripheral malaria infections, 32 had placental malaria infections, and 42 had infections in one or both compartments. No cord blood malaria was identified in this study. Table 1 cross-references malaria infections by compartment and diagnostic method.
Table 2 shows the characteristics of study enrollees classified by maternal peripheral malaria. All women were HIV negative. Women with and without peripheral malaria infections were similar in terms of age and anthropometric measures. However, malaria-infected women were younger, of lower parity, and tended to have fewer antenatal clinic visits (Table 2). Socioeconomic status, as measured by educational level, house construction, and employment status, did not differ between the groups (Table 2). When women were alternatively classified by placental malaria status, there were again no significant differences in study participant characteristics (data not shown). However, when reclassified according to malaria infection in any compartment (maternal and/or placental), malaria-infected women had significantly lower BMI (23.4 ± 4.0 vs. 24.8 ± 2.4; p = 0.02) and fewer antenatal clinic visits (4.8 ± 1.8 vs. 5.7 ± 2.5; p = 0.045); no other characteristics differed between the groups (data not shown).
Data on the effects of malaria on maternal and foetal markers of haematological status, inflammation, and chronic hypoxia are presented in two tables: Table 3 presents the effects of maternal peripheral malaria on the markers, and Table 4 presents the effects of placental malaria.
Table 4 Mann-Whitney U test of effect of placental malaria on markers of neonatal haematological, inflammatory, and hypoxic status
Infected median (IQRa) n (total n: 32) Uninfected median (IQR) n (total n: 58)
Cord
Hb (g/dL) 15.4 (13.7 – 16.8) 25 16.0 (15.0 – 17.1) 45 p = 0.128
sTfR (μg/ml) 6.7 (5.1 – 9.9) 32 7.7 (4.3 – 9.6) 58 p = 0.765
ferritin (ng/ml) 129.0 (61.9 – 248.6) 32 101.0 (41.2 – 210.8) 57 p = 0.184
TNF-α (pg/ml) 1.7 (0.0 – 5.6) 30 2.7 (0.0 – 15.9) 57 p = 0.552
TGF-β (ng/ml) 19.5 (15.6 – 30.3) 16 26.8 (17.9 – 31.8) 31 p = 0.080
CRP (μg/ml) 1.5 (0.7 – 1.6) 31 1.0 (0.0–1.6) 58 p = 0.289
Epo (mU/L) 25.4 (13.4 – 58.1) 32 27.5 (12.5 – 45.1) 58 p = 0.940
Cortisol (μg/ml) 421.0 (283.5 – 538.5) 32 511.0 (324.0 – 677.5) 53 p = 0.425
Placenta
Cortisol (μg/ml) 191.0 (155.5 – 251.0) 29 172.0 (132.0 – 248.5) 53 p = 0.357
CRH (pg/ml) 90.0 (42.5 – 126.4) 24 72.3 (41.5 – 116.3) 43 p = 0.421
aIQR: Inter-quartile range
The effect of malaria on haematological status differed markedly between women and neonates. Compared to their uninfected counterparts, women characterized by any of the three definitions of malaria infection tended to have significantly lower maternal Hb concentrations (maternal: Table 3; placental: p = 0.017; any: p < 0.001). Unlike maternal haematological status, neonatal characteristics were protected from the effects of maternal and placental malaria. Mean cord Hb levels did not differ between malaria-infected and uninfected women (Tables 3 and 4). In addition, malarial disease severity, including peripheral and placental parasite density, stage of disease, and deposition of fibrin and malaria pigment in placental monocytes, had no effect on neonatal Hb (data not shown). Neither cord nor maternal sTfR levels were affected by malaria infection in this study (Tables 3 and 4).
Markers of inflammation were also altered in malaria-infected women, who had significantly elevated CRP and TNF-α (maternal: Table 3; placental: CRP (p < 0.001), TNF-α (p = 0.022); any: CRP (p < 0.001), TNF-α (p = 0.004)) and lower TGF-β, although this result was not significant after Bonferroni's adjustment (maternal: Table 3; placental: NS; any: p = 0.010). There was no effect of maternal malaria on foetal TNF-α or CRP (Table 3). TGF-β was significantly decreased in the cord blood of neonates with peripheral malaria-infected mothers, although this result was not significant after Bonferroni's adjustment (Table 3). Among babies of malaria-infected women, neither increased maternal or placental parasite density nor markers of disease severity were associated with increases in measures of neonatal inflammation (data not shown). However, elevated maternal CRP in women with peripheral malaria infections was associated with significantly increased cord CRP (r = 0.48, p = 0.010) and decreased cord TGF-β (r = -0.60, p = 0.019), suggesting an association between maternal and foetal immune activation.
Malaria-infected women had significantly higher ferritin levels than uninfected women (maternal: Table 3; placental: p < 0.001; any: p < 0.001). Cord ferritin was significantly increased in the cord blood of neonates with peripheral malaria-infected mothers, although this result was not significant after Bonferroni's adjustment (Table 2). In addition, cord ferritin was correlated with increasing maternal parasite load in malaria-infected women (r = 0.474, p = 0.009). The notion that ferritin levels in this group reflected malaria-related inflammation is further supported by the inverse correlation between cord ferritin and cord TGF-β (r = -.593, p = 0.020). Additionally, primigravids with placental malaria infections, who tend to experience its inflammatory effects more acutely than multigravids, had significantly higher cord ferritin levels than did multigravid malaria-infected women (p = 0.021). However, despite the indications of immune activation in neonates of malaria-infected women, none of the cord immune markers were correlated with cord Hb or sTfR (data not shown).
Although women characterized by any of the three definitions of malaria infection tended to have significantly higher Epo levels than uninfected women (maternal: Table 3; placental: NS; any: p = 0.004), none of the proxies for foetal hypoxia examined in the study (cord Epo, placental and cord cortisol, and placental CRH) were increased in the neonates of malaria-infected women (Tables 3 and 4). Among malaria-infected women, increased maternal and placental parasitaemia and markers of disease severity were also not associated with increased levels of these factors (data not shown).
Neonatal haematological, immune and hypoxic markers were then examined in relation to poor birth outcomes that have been associated with maternal malaria, particularly low birthweight and preterm delivery. In this study, neither the presence nor the severity of maternal or placental malaria was associated with decreased birthweight or gestational age, perhaps due to the limited sample size (data not shown). Unsurprisingly, therefore, neither decreased maternal nor cord Hb was significantly associated with poorer birth outcomes in malaria-infected women (data not shown), although cord sTfR levels, which were not different between neonates of malaria-infected and uninfected women (Tables 3 and 4), were inversely associated with birthweight in women with maternal, but not placental, malaria (r = -.452, p = 0.014). In neonates of women with both maternal and placental malaria, elevated cord ferritin was correlated with lower birthweights (maternal: r = -0.427, p = 0.021; placental: r = -0.374, p = 0.035), and in those of women with maternal malaria, with shorter gestations as well (r = -0.432, p = 0.024). This association was not due to inflammation alone, since there was no association with cord CRP or TGF-β, nor to foetal haematopoiesis alone, since there was no association with cord Epo (data not shown). However, the data suggest that a combination of the haematological and inflammatory processes might be associated with poor birth outcomes commonly found following malaria infection during pregnancy.
Discussion
In this study, the haematological status of neonates of mothers with malaria parasitaemia at delivery was not significantly impacted. Despite the effect of malaria on maternal Hb, which was expectedly lower, neonatal Hb levels were not decreased in the cord blood of malaria-infected compared to uninfected women. In addition, cord Hb values were not correlated to maternal Hb concentration in either group or severity of malaria, as described in the Results section. Although the sample size was relatively small, these results resemble more closely the results from Western studies, which tend to find unaffected cord values, than the studies of malaria- and non-malaria-exposed mother-neonate pairs in developing countries, as reviewed by Brabin [4]. It is unclear why, but the high cord haemoglobin levels observed in this study may relate to the high rates of antimalarial usage and iron supplementation in Malawian women (Table 2), which might provide a protective effect. The maintenance of cord Hb levels despite the presence of maternal anaemia and malaria suggests that the foetus has developed mechanisms to preferentially obtain sufficient iron and produce adequate amounts of red cells. Iron is transported unidirectionally from mother to foetus across a concentration gradient [31], and thus stores should be preferentially preserved in the foetus.
Ferritin is an iron storage protein that is an acute phase reactant in the presence of infection. While cord ferritin decreases in foetuses of anaemic mothers without malaria, it appears to increase in foetuses of mothers with increasing parasitaemia. Increased ferritin in the presence of inflammation has been suggested to reflect an evolutionary strategy by which the body removes bioavailable iron from the presence of parasites [32]. In foetuses of mothers with malaria, it may have an additional function: to help sequester iron obtained from the mother. Interestingly, in this study, elevated cord ferritin levels best predicted poor birth outcomes, including shorter gestation and lower birth weight. It is unclear why this association was found, particularly since no association was found in this study between birth outcomes and malaria infection or maternal anaemia, two well-established risk factors. This suggests that increased cord ferritin might be a result of the same foetal stresses that cause poor birth outcomes.
Many researchers have assumed that foetal hypoxia, rather than inflammation, mediates the effect of malaria and maternal anaemia on birth outcomes, particularly low birth weight. Despite the induction of foetal hypoxia by maternal anaemia [20,33] and the decrease in umbilical blood flow in pregnant women with malaria [34], it has not been rigorously demonstrated that chronic low-grade hypoxia causes poor birth outcomes, and it is unclear how severe anaemia must be to induce foetal hypoxia. Foetal hypoxia is difficult to quantify in nonwestern settings. In developed countries, blood gases are collected as soon as the foetal scalp is discernable exiting the birth canal. This technique is not feasible in hospitals in developing countries, but several biochemical measures in cord blood may offer proxies for foetal hypoxia. In particular, Epo, the hormone primarily responsible for regulation of erythropoiesis, is stimulated by both anaemia and hypoxia [35] and is a good marker of chronic foetal hypoxia [19,36] because Epo does not cross the placenta. Several studies have reported both adequate [37-39] and inadequate [40,41] Epo production for the level of anaemia in malaria-infected children and adults, and this question requires further investigation. In this study, maternal Epo was elevated in malaria-infected women (Table 3), but cord Epo levels were not affected by either maternal or placental malaria infections (Tables 3 and 4).
Allen [20] hypothesized that foetal stress hormones (cortisol and CRH) might mediate the link between maternal anaemia and poor birth outcomes, via the effect of anaemia-induced foetal hypoxia on these hormones. In this study, cord cortisol and placental cortisol and CRH were neither elevated in association with maternal malaria infection nor associated with poor birth outcomes. This may reflect the inadequacy of these factors as markers for chronic foetal hypoxia, either in general or in cord blood, or may simply be a result of the small sample size. On the other hand, hypoxia in the foetus may be so fleeting, as the foetus downregulates its growth in relation to available oxygen levels, that it does not trigger lasting detectable markers. This response is seen in sheep, where pregnant females with experimentally-induced anaemia evince a compensatory decrease in foetal growth but no actual evidence of hypoxia [42].
Conclusion
In summary, cord haemoglobin levels were unchanged by maternal malaria and anaemia during pregnancy in this population. However, maternal parasitaemia induced a foetal inflammatory response, specifically lower TGF-β and higher ferritin levels. Elevated cord ferritin in turn was associated with lower birthweights and shorter gestations.
Authors' contributions
ETA was responsible for conception and design of the study; IRB approval; conducting assays; analysis and interpretation of data; drafting the paper; and revising it for publication. JK and DDK performed many of the assays and analyses of these assays. VM designed the larger study and conducted much of the data analysis of that study that formed the foundation of this one. ET and VLM supervised patient enrollment and acquisition of data and aided in its submission to the IRB. MEM and SRM were central to the conception and design of the study, the analysis and interpretation of data, and the revision of the paper for publication. SRM was involved in the conception and design of the study; IRB approval; analysis and interpretation of data; drafting the paper; and revising the paper for publication.
Acknowledgements
Funding for this study was provided by The Wenner-Gren Foundation for Anthropological Research (ETA), the Anthropology Department of the University of Michigan (ETA), the Rackham School of Graduate Studies at the University of Michigan (ETA), the Wellcome Trust Career Development Fellowship (SJR; ref 046012), and the NIH (SRM; AI 49084).
==== Refs
Brabin BJ An analysis of malaria in pregnancy in Africa Bull World Health Organ 1983 61 1005 1016 6370484
Menendez C Ordi J Ismail MR Ventura PJ Aponte JJ Kahigwa E Font F Alonso PL The impact of placental malaria on gestational age and birth weight J Infect Dis 2000 181 1740 1745 10823776 10.1086/315449
Menendez C Fleming AF Alonso PL Malaria-related anaemia Parasitol Today 2000 16 469 476 11063857 10.1016/S0169-4758(00)01774-9
Brabin B Foetal anaemia in malarious areas: its causes and significance Ann Trop Paediatr 1992 12 303 310 1280047
Erdem A Erdem M Arslan M Yazici G Eskandari R Himmetoglu O The effect of maternal anaemia and iron deficiency on fetal erythropoiesis: comparison between serum erythropoietin, haemoglobin and ferritin levels in mothers and newborns J Matern Fetal Neonatal Med 2002 11 329 332 12389675
Hokama T Takenaka S Hirayama K Yara A Yoshida K Itokazu K Kinjho R Yabu E Iron status of newborns born to iron deficient anaemic mothers J Trop Pediatr 1996 42 75 77 8984218 10.1093/tropej/42.2.75
Agarwal KN Krishna M Singla PN Jain A Effects of hematinic and protein supplementation on the feto-placental unit in maternal anaemia Indian Pediatr 1984 21 765 771 6526483
Tracer DP Reproductive and socio-economic correlates of maternal haemoglobin levels in a rural area of Papua New Guinea Trop Med Int Health 1997 2 513 518 9236817 10.1046/j.1365-3156.1997.d01-327.x
Bhargava M Iyer PU Kumar R Ramji S Kapani V Bhargava SK Relationship of maternal serum ferritin with foetal serum ferritin, birth weight and gestation J Trop Pediatr 1991 37 149 152 1960769
Ballew C Haas JD Haematologic evidence of foetal hypoxia among newborn infants at high altitude in Bolivia Am J Obstet Gynecol 1986 155 166 169 2425621
Abdalla SH Haematopoiesis in human malaria Blood Cells 1990 16 401 416 discussion 417-409 2257320
Huddle JM Gibson RS Cullinan TR The impact of malarial infection and diet on the anaemia status of rural pregnant Malawian women Eur J Clin Nutr 1999 53 792 801 10556986 10.1038/sj.ejcn.1600851
Rogerson SJ Pollina E Getachew A Tadesse E Lema VM Molyneux ME Placental monocyte infiltrates in response to Plasmodium falciparum malaria infection and their association with adverse pregnancy outcomes Am J Trop Med Hyg 2003 68 115 119 12556159
Fried M Muga RO Misore AO Duffy PE Malaria elicits type 1 cytokines in the human placenta: IFN-gamma and TNF-alpha associated with pregnancy outcomes J Immunol 1998 160 2523 2530 9498798
Rogerson SJ Brown HC Pollina E Abrams ET Tadesse E Lema VM Molyneux ME Placental tumor necrosis factor alpha but not gamma interferon is associated with placental malaria and low birth weight in Malawian women Infect Immun 2003 71 267 270 12496175 10.1128/IAI.71.1.267-270.2003
Moormann AM Sullivan AD Rochford RA Chensue SW Bock PJ Nyirenda T Meshnick SR Malaria and pregnancy: placental cytokine expression and its relationship to intrauterine growth retardation J Infect Dis 1999 180 1987 1993 10558956 10.1086/315135
Fievet N Ringwald P Bickii J Dubois B Maubert B Le Hesran JY Cot M Deloron P Malaria cellular immune responses in neonates from Cameroon Parasite Immunol 1996 18 483 490 9226685 10.1046/j.1365-3024.1996.d01-19.x
King CL Malhotra I Wamachi A Kioko J Mungai P Wahab SA Koech D Zimmerman P Ouma J Kazura JW Acquired immune responses to Plasmodium falciparum merozoite surface protein-1 in the human foetus J Immunol 2002 168 356 364 11751981
Buescher U Hertwig K Wolf C Dudenhausen JW Erythropoietin in amniotic fluid as a marker of chronic foetal hypoxia Int J Gynaecol Obstet 1998 60 257 263 9544710 10.1016/S0020-7292(98)00011-3
Allen LH Biological mechanisms that might underlie iron's effects on foetal growth and preterm birth J Nutr 2001 131 581S 589S 11160591
Pacifici GM Nottoli R Placental transfer of drugs administered to the mother Clin Pharmacokinet 1995 28 235 269 7758253
Zanjani ED Pixley JS Slotnick N MacKintosh FR Ekhterae D Clemons G Erythropoietin does not cross the placenta into the foetus Pathobiology 1993 61 211 215 8216843
Starreveld JS Kroos MJ Van Suijlen JD Verrijt CE Van Eijk HG Van Dijk JP Ferritin in cultured human cytotrophoblasts: synthesis and subunit distribution Placenta 1995 16 383 395 7567801 10.1016/0143-4004(95)90096-9
Zaretsky MV Alexander JM Byrd W Bawdon RE Transfer of inflammatory cytokines across the placenta Obstet Gynecol 2004 103 546 550 14990420
Gutteberg TJ Askvik K Jorgensen T Serum lactoferrin and C-reactive protein in mother and newborn after preterm rupture of membranes Acta Obstet Gynecol Scand 1986 65 203 205 3488635
Yang K Placental 11 beta-hydroxysteroid dehydrogenase: barrier to maternal glucocorticoids Rev Reprod 1997 2 129 132 9414475 10.1530/ror.0.0020129
Mwapasa V Rogerson SJ Molyneux ME Abrams ET Kamwendo DD Lema VM Tadesse E Chaluluka E Wilson PE Meshnick SR The effect of Plasmodium falciparum malaria on peripheral and placental HIV-1 RNA concentrations in pregnant Malawian women Aids 2004 18 1051 1059 15096809 10.1097/00002030-200404300-00014
Ballard JL Khoury KC Wedig K Wang L Eilers-Walsman BL Lipp R New Ballard Score, expanded to include extremely premature infants J Pediatr 1991 119 417 423 1880657
Rogerson SJ Chaluluka E Kanjala M Mkundika P Mhango C Molyneux ME Intermittent sulfadoxine-pyrimethamine in pregnancy: effectiveness against malaria morbidity in Blantyre, Malawi, in 1997–99 Trans R Soc Trop Med Hyg 2000 94 549 553 11132387 10.1016/S0035-9203(00)90083-X
Sokal RR Rohlf FJ Biometry 1995 New York: W. H. Freeman and Company
Fletcher J Suter PE The transport of iron by the human placenta Clin Sci 1969 36 209 220 5772100
Weinberg ED Nutritional immunity. Host's attempt to withold iron from microbial invaders Jama 1975 231 39 41 1243565 10.1001/jama.231.1.39
Rondo PH Tomkins AM Maternal iron status and intrauterine growth retardation Trans R Soc Trop Med Hyg 1999 93 423 426 10674095 10.1016/S0035-9203(99)90144-X
Arbeille P Carles G Bousquet F Body G Lansac J Foetal cerebral and umbilical artery blood flow changes during pregnancy complicated by malaria J Ultrasound Med 1998 17 223 229 9544605
Vora M Gruslin A Erythropoietin in obstetrics Obstet Gynecol Surv 1998 53 500 508 9702790 10.1097/00006254-199808000-00023
Ostlund E Lindholm H Hemsen A Fried G Foetal erythropoietin and endothelin-1: relation to hypoxia and intrauterine growth retardation Acta Obstet Gynecol Scand 2000 79 276 282 10746842 10.1034/j.1600-0412.2000.079004276.x
Burchard GD Radloff P Philipps J Nkeyi M Knobloch J Kremsner PG Increased erythropoietin production in children with severe malarial anaemia Am J Trop Med Hyg 1995 53 547 551 7485716
Nussenblatt V Mukasa G Metzger A Ndeezi G Garrett E Semba RD Anaemia and interleukin-10, tumor necrosis factor alpha, and erythropoietin levels among children with acute, uncomplicated Plasmodium falciparum malaria Clin Diagn Lab Immunol 2001 8 1164 1170 11687458 10.1128/CDLI.8.6.1164-1170.2001
Verhoef H West CE Kraaijenhagen R Nzyuko SM King R Mbandi MM van Laatum S Hogervorst R Schep C Kok FJ Malarial anaemia leads to adequately increased erythropoiesis in asymptomatic Kenyan children Blood 2002 100 3489 3494 12393621 10.1182/blood-2001-12-0228
el Hassan AM Saeed AM Fandrey J Jelkmann W Decreased erythropoietin response in Plasmodium falciparum malaria-associated anaemia Eur J Haematol 1997 59 299 304 9414641
Vedovato M De Paoli Vitali E Dapporto M Salvatorelli G Defective erythropoietin production in the anaemia of malaria Nephrol Dial Transplant 1999 14 1043 1044 10328513 10.1093/ndt/14.4.1043
Mostello D Chalk C Khoury J Mack CE Siddiqi TA Clark KE Chronic anaemia in pregnant ewes: maternal and foetal effects Am J Physiol 1991 261 R1075 1083 1951756
|
16159390
|
PMC1232865
|
CC BY
|
2021-01-04 16:36:35
|
no
|
Mol Cancer. 2005 Sep 13; 4:36
|
latin-1
|
Mol Cancer
| 2,005 |
10.1186/1476-4598-4-36
|
oa_comm
|
==== Front
Popul Health MetrPopulation Health Metrics1478-7954BioMed Central London 1478-7954-3-91610518010.1186/1478-7954-3-9ResearchCollection of population-based cancer staging information in Western Australia – a feasibility study Threlfall Timothy [email protected] Jana [email protected] Padabphet [email protected] Jane [email protected] Paul [email protected] Harry [email protected] Lin [email protected] Western Australian Cancer Registry, Department of Health, Government of Western Australia, Perth, Australia2 School of Population Health, University of Western Australia, Perth, Australia3 Western Australian Clinical Oncology Group, The Cancer Council of Western Australia, Perth, Australia2005 17 8 2005 3 9 9 28 6 2005 17 8 2005 Copyright © 2005 Threlfall et al; licensee BioMed Central Ltd.2005Threlfall 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
Routine data from cancer registries often lack information on stage of cancer, limiting their use. This study aimed to determine whether or not it is feasible to add cancer staging data to the routine data collections of a population-based Western Australian Cancer Registry (WACR).
Methods
For each of the five most common cancer types (prostate, colorectal, melanoma, breast and lung cancers), 60 cases were selected for staging. For the 15 next most common cancer types, 20 cases were selected. Four sources for collecting staging data were used in the following order: the WACR, the hospital based cancer registries (HBCRs), hospital medical records, and letters to treating doctors. If the case was unable to be fully staged, due to lack of information on regional lymph node invasion or distant metastases, we made the following assumptions. Cases which had data available for tumour (T) and regional lymph nodes (N), but no assessment of distant metastasis (MX) were assumed to have no distant metastases (M0). Cases which had data for T and M, but no assessment of regional nodal involvement (NX) were assumed to have no regional nodal involvement (N0).
Results
The main focus of this project was the process of collecting staging data, and not the outcomes. For ovary, cervix and uterus cancers the existence of a HBCR increased the stageable proportion of cases so that staging data for these cancers could be incorporated into the WACR immediately. Breast and colorectal cancer could also be staged with adequate completeness if it were assumed that MX = M0. Similarly, melanoma and prostate cancer could be staged adequately if it were assumed that NX = N0 and MX = M0. Some cases of stomach, lung, pancreas, thyroid, testis and kidney cancers could be staged, but additional clinical input – on pathology request forms, for example – would be required to achieve useable levels of completeness. For the remaining cancer types either staging is widely regarded as not relevant, and no generally-accepted system exists, or an acceptable level of completeness is not achievable.
Conclusion
Adding stage to routinely collected information in a cancer registry is possible for many cancer types, particularly if the assumptions regarding missing data are found to be acceptable or if the guidelines for MX = M0 asumptions are clarified. These findings should be generalizable to most cancer registries in developed countries, if hospital-based cancer registries or other specialized databases are accessible.
==== Body
Background
Cancer staging information is of fundamental importance at both population and individual levels. For the individual, it facilitates provision of appropriate patient care, enables appropriate selection of treatment for individual cases, can be used to explain variability in treatment outcomes, and can contribute to helping an individual patient and their family to better understand the clinical condition and prognosis. At the population level, staging data can guide the development and evaluation of health promotion and treatment programs, can facilitate more effective resource allocation depending on the relative proportions of "early" as opposed to "late" cases and can be used to stratify survival analyses to improve comparison of results of different groups such as geographic areas [1-3]. The lack of staging data has been recognised by clinicians and public health professionals alike as an important limitation of cancer registry data.
Different staging systems exist and they have different rules and guidelines. Commonly used staging systems include those of the American Joint Committee on Cancer (AJCC), International Union Against Cancer (UICC) and International Federation of Gynaecology and Obstetrics (FIGO) which all rely upon characteristics of primary tumour, nodes and metastases as a basis. The Ann Arbor and Dukes classifications also use similar definitions and principles [4,5].
The TNM system depends on the combination of different characteristics of the tumour. T describes the primary tumour size and/or extent, N describes the presence or absence of regional lymph node metastasis and M describes the presence or absence of distant metastasis. Each of these components is divided into numerical subsets (T0 – T4, N0 – N3, M0 – M1) which describe how advanced the malignancy is. The definitions of these subsets are specific for each tumour and are delineated in a handbook [4]. Depending on the specific combination of T, N and M, an individual cancer will be assigned to a "stage".
For most cancers, the staging of the tumour depends only on TNM. However, some tumours also require additional information for staging (for example, serum tumour markers for testicular cancer) [4]. Additional prognostic factors may be increasingly included in the delineation of the TNM stage grouping. The TNM system is flexible and accepted worldwide for patient care, and has been validated as being relevant to the clinical practice of oncology [6].
Western Australia is the largest state of Australia. With an area of more than 2 500 000 sq km, a 12 500 km coastline and spanning 2 400 km from north to south, it occupies a third of the continent. The total population of Western Australia is just under 1.5 million people of whom 1.2 million live in and around the capital city of Perth.
The Western Australian Cancer Registry (WACR) is a population-based cancer registry, which was established in 1981, based on the mandatory reporting of cancers by pathologists, haematologists and radiation oncologists [7]. The WACR routinely collects data relating to tumour location, type, basis and date of diagnosis, and grade, together with demographic information and some possible sources of further information. At present the WACR does not routinely register information on cancer stage. Recognising the need for these data, this project aimed to determine the feasibility, in terms of ascertainment and effort required, of adding staging data to the WACR.
There are four major public hospitals in Perth and three major private hospitals. Each of the public hospitals has a Hospital Based Cancer Registry (HBCR) which collects information on a limited range of cancer types, including information on staging at the time of diagnosis.
Methods
The project aimed to collect staging data on a selection of cancer cases from among those notified to the WACR, including cases notified in the past (retrospective data collection) and in the present (prospective data collection). The retrospective cases were those diagnosed in 1998 (300) and in January – June 2002 (150); the prospective cases (150) were those diagnosed after June 2002. The cancer types were selected on the basis of the 20 most common incident cancers in 2000 (apart from non-melanocytic skin cancers, which are not collected by WACR). For each of the five most common cancer types (prostate, colorectal, melanoma, breast and lung cancers) 60 cases were randomly selected for staging and for the remaining 15 cancer types, 20 cases were selected. Approval for access to medical records was requested from each hospital in which selected cases had been treated.
There were 4 sources for collecting data for the staging: the WACR, the HBCRs, hospital medical records, and responses to enquiry letters to treating doctors. For most cancers, this was the order in which the sources were approached. The first step involved reviewing the WACR files and extracting available staging data from pathology reports. If the cancer could not be fully staged from the WACR files, the HBCRs were approached for any staging information they held. If full staging information was not available from WACR or HBCRs, a project officer reviewed casenotes at both private and public hospitals. For reasons related to costs, medical records were not searched at country hospitals and small city hospitals. If the cancer was still unable to be staged, a letter was sent to the treating clinician requesting staging information.
For some cancers, specific local circumstances meant it was more efficient to undertake the steps in a different order, or to omit some steps. Gynaecological cancers (cervix, uterus and ovary), are to a large extent treated in one public hospital with an active HBCR. The FIGO summary stage for all gynaecological cancers was generally the only information used for these cancers. For colorectal cancer, information was obtained initially from the HBCRs and also obtained from the WA Research Tissue Network, which collects statewide data (including stage) on certain cancers. Cases which could not be staged using these resources were then researched using other sources as necessary. For the lymphohaematopoietic malignancies, for which the TNM system is not appropriate, and for brain cancers, WACR files only were reviewed. For melanoma, most cases do not require hospital admission so HBCRs and hospital casenote reviews were not done.
It should be noted that the stage at diagnosis commonly includes data collected over a period of up to 4 months [8] although the WACR uses 3 months as a routine.
We calculated the proportion of cases that could be staged after each step of the staging process overall and stratified by prospective and retrospective groupings.
In order to definitively stage many cancers, considerable clinical investigation is required. For example, lymph nodes need to have been dissected or biopsied, or extensive searches for metastases need to have been undertaken. These further investigations have associated costs and risks and may not always be clinically warranted. For example, if a melanoma is found to be level 1, and there is no clinical evidence of spread, investigations such as chest X-rays and bone scans are unlikely to be performed. In addition, some cases do undergo further investigations but the results are held in private rooms, or medical records do not contain any negative information such as the absence of metastases. In addition, doctors may not have responded to our letters asking for information on stage. For many of the cases in this study, therefore, it was not possible to stage the cancer definitively because of the lack of information on regional nodal status or the presence or absence of distant metastases. Many of these cases are most likely to be early stage cancers.
The data for those cases that were not stageable were further examined, and two different assumptions were applied. The first assumption was applied to cases which had data for T and N but which had no assessment of distant metastasis (MX). The assumption made was that MX was equivalent to M0 (no distant metastases). This is summarized as MX = M0 in the tables.
The second assumption was applied to cases with data for T and M, but with no assessment of regional nodal involvement (NX). The assumption made was that NX was equivalent to N0 (no regional nodal involvement). This is summarized as NX = N0 in the tables.
The number of cases that could be staged if both assumptions were made for the same case (ie NX = N0 and MX = M0) was also calculated. All original data (without assumptions) was stored separately.
The costs of collecting staging data were also estimated. For each cancer type, actual times from the study were extrapolated to costs based on a whole year's collection of staging data. Tasks contributing to this time included: reviewing pathology reports; looking at electronic files at WACR; examining medical case notes; writing enquiry letters and reviewing replies. Added to this were general costs including travel, liaison with data holders, training, leave, general office duties, and delays due to competing priorities of hospital and other non-WACR staff.
Results
The focus of this project was the process of collecting staging data, and not the actual stages of the cancers.
The main finding from this feasibility study is that staging completeness is very dependent on cancer type (Table 1). Each type has its own issues and complications. However, in order to summarize the results we have grouped some types together as follows:
Table 1 Percentage of cases staged, by cancer type and process used. Percentages shown are cumulative, beginning from the left.
Cancer type Staged from WACR data alone (%) Staged from WACR and HBCRs, WARTN (%) Staged after all completed steps (no assumptions) (%) Staged after all steps with assumption/s (%)
Group A: Could be staged now
Ovary 60^ 100 100 100
Cervix 16^ 95 100 100
Uterus 50^ 85 95 95
Group B: Could be staged now, making MX = M0 assumption
Breast 0 12 65 95
Colorectal 12^ 53 80 92
Group C: Could be staged now, with NX = N0 and MX = M0 assumptions
Melanoma 0 (0) 57 100
Prostate 2 5 34 97
Group D: Could be started now with MX = M0, but long term collection requires system changes
Stomach** 25 25 70 95
Lung 18 38 76 86
Pancreas 45 45 70 80
Thyroid** 10 10 47 79
Testis** 10 10 75 75
Kidney** 15 20 65 70
Group E: Staging not feasible at present
Oesophagus** 0 0 50 65
Bladder** 0 0 40 55
Lip 0 (0) 37 42
Lymphoma 44 (44) 44 44
Myeloma 0 (0) 0 0
Leukaemia 0 (0) 0 0
Brain 0 (0) 0 0
^ These numbers could have been higher as the external databases were searched first, and WACR later searched only for incomplete cases.
** Only one HBCR currently collects data on these cancers except bladder, for which two HBCRs are collecting data.
() Numbers in parentheses indicate that the additional data source/s indicated by the column header, was/were not accessed as they were either not applicable to the cancer type, or research suggested the additional effort would be unrewarding.
• Group A: cancers for which virtually complete staging data could be obtained relatively easily. Relatively complete information on stage of these cancers was available with current systems.
• Group B: cancers for which relatively high proportions of cancers could be fully staged, and for which the assumption MX = M0 allows almost complete staging. If it was considered reasonable to apply this assumption, relatively complete information on stage would be available on these cancers.
• Group C; cancers which can almost all be staged making the assumptions NX = N0 and MX = M0. If it was considered reasonable to apply these assumptions, relatively complete staging information could be collected.
• Group D; cancers for which it is more risky to apply the assumptions. Accurate collection requires system changes in order to obtain better information on stage.
• Group E; cancers for which staging is not feasible at present.
Group A cancers (Table 1) consisted of the gynaecological cancers: ovary, cervix and uterus. The use of the HBCR at the gynaecological hospital was clearly crucial for staging these cancers, increasing the stageable proportion of cases from 60% to 100% for ovarian cancer and from 16% to 95% for cervical cancer. For these cancers there was no need to apply any assumptions. Relatively complete staging data could be collected easily.
Group B cancers consisted of breast and colorectal cancer. For these two cancers, a reasonable number of cases could be staged fully using the standard 4 steps. For colorectal cancer, 80% of cases could be staged after all four steps were completed, with this increasing to 92% if the assumption MX = M0 was applied. For breast cancer 65% of the cases could be staged after all four steps were completed, increasing to 95% when applying the assumption MX = M0.
Group C cancers consisted of melanoma and prostate cancer. Only 34% of prostate cancer cases and 0% of melanoma cases could be staged using only the WACR, mainly because of lack of information on nodal status and metastases. However, if both assumptions of NX = N0 and MX = M0 were applied this increased to 100% of prostate cancers and 97% of melanomas.
Group D cancers include cancers of the stomach, lung, pancreas, thyroid, testis and kidney. Cancer types in Group D would require additional clinical input to achieve adequate proportions of staged cancers. Making the assumption of MX = M0, a relatively high proportion of cases may be able to be staged. However, the acceptability of the assumptions differs among cancers.
Cancer types in Group E are not able to be staged at present. After all four steps were completed only 50%, 40% and 37% of cases could be staged, for cancers of the oesophagus bladder and lip, respectively. These levels of completeness were still inadequate even after applying the assumption MX = M0. No myeloma or leukaemia cases could be staged at only step 1, and the only lymphoma cases that could be staged were Stage 4. The amount of clinical information required to stage these cancers goes far beyond the pathological details available. There is no accepted staging system for brain cancer.
Costs were estimated assuming that the HBCRs continued to operate. In this case, collection of staging data on Group A cancers would cost approximately 0.1 of a full-time equivalent staff member. Costs of collecting staging data for Groups A and B would require about one half-time staff member and to collect staging data on Groups A, B and C would require about one full-time staff member. This is in relation to a population of 1.5 million.
Discussion
The main finding from this study is that collection of adequate staging data by population-based cancer registries depends primarily on the type of cancer and the existence of HBCRs or other specialized registries which already collect such data. Because HBCRs already collect high quality staging information in Western Australia, their operation is vital to the efficiency and cost-effectiveness of population-based collection of staging data. As can be seen from the results of this study, in cases where HBCRs covered a large proportion of the cancers seen in WA (for example the gynaecological cancers, or colorectal cancer), the proportion of cases which could be staged was high. In other cases, such as melanoma, for which no HBCR collects staging information, the proportion of cases which could be staged tended to be low. The greater the coverage of HBCRs in terms of cancer types and hospitals, the better the stage collection of the population-based registry can be. While this is cost-efficient for the population based registry, costs have to be borne by the HBCRs.
In this study, we had a short timeline which meant that the data we were able to access from the HBCRs was probably less complete than would be available over a longer period. On the other hand, the fact that we did not collect information from small country hospitals means that we probably slightly overestimated the proportion of cancers which could be staged.
One of the problems in attempting to collect staging data for a variety of different cancer types was that population-based cancer registries such as the WACR usually rely on pathology reports, and do not routinely receive non-pathology reports. These include radiological reports such as CT scans, X-rays and PET scans which contain information on N and M. Other information not routinely supplied to registries may be crucial to staging, such as hormonal assays for testicular cancer, negative lymph node biopsies for breast cancer, and haemoglobin level and cell counts in leukaemia.
Related to this was the necessity to make assumptions about MX and NX. The acceptability of these assumptions needs to be ascertained and is different for the different cancer sites. For example, for melanomas of Clark level I [9], or low Breslow thickness [10], it is probably quite reasonable to assume that NX = N0 and MX = M0. However, for stomach and pancreatic cancer, the viability of the MX = M0 assumption is questionable as these cancers are often at a late stage when diagnosed. This means that cancer sites which had similar proportions of stageable cancers without assumptions were placed in different categories depending on how clinical acceptable the assumptions were. For example, only 54% of the melanomas could be staged without assumptions, and 100% with assumptions and clinically it was thought that the assumptions were reasonable. However, 70% of the stomach cancers could be staged without assumptions, and 95% with assumptions but clinically it was not thought that the assumptions were reasonable.
Where it is not thought appropriate to apply the MX = M0 assumption to all unstaged cancer cases, an alternative strategy would be to work with clinicians in order to develop rules about for which cases it would be appropriate to apply this assumption. A separate study suggests that it is safe to make the MX = M0 assumption in 90% of breast cancer cases with T1, T2 or T3 and either N0 or N1 (personal communication, Padabphet Boutdara) and about 85% of similar colorectal cancer cases [11].
A more difficult problem occurred when the cancer was diagnosed on a biopsy specimen. In these cases, even the information on tumour size (T) was often unable to be ascertained.
Increasing the proportion of cancers which could be staged would require a greater level of input from clinicians than currently exists. For instance, the use of synoptic structured pathology reports would assist the staging of these cancers, by providing a consistent approach to staging, as well as an easy format for the clinicians to provide this information to the cancer registry.
In order to guide our comments on the feasibility of various potential recommendations we interviewed specialist clinicians interested in specific cancer types, to canvass their views on issues relating to staging data collection. Some of the issues raised by clinicians were confidentiality and privacy concerns, and access to and "ownership" of data. In addition, there was disagreement among the clinicians as to whether it was the responsibility of clinicians to do the staging, or whether that should be done by registry staff.
There are several other problems with adding staging to a population based registry. If a formal screening program is introduced there may be better recording of data and less missing information on stage over time. In addition it is important to consider whether staging in a population-based cancer registry is worthwhile if reasonable levels of completeness cannot be achieved. For example, it may be hard to justify collection of data for kidney, testicular, thyroid or pancreatic cancer as 20% – 30% of cases in a statistical analysis would not have staging information.
Conclusion
Adding stage to a population-based cancer registry is highly dependent on the type of the cancer. However, routinely collected staging is possible for many cancer types with the co-operation of HBCRs and the assumptions of negative nodal or metastatic status for some types. These findings should be generalizable to most cancer registries in developed countries, if hospital-based cancer registries or other specialized databases are accessible.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
TT participated in the design of the study, managed the project staff, oversaw the analysis and project report, and assisted with drafting the manuscript.
JW was responsible for the co-ordination of data collection, provided the results for the project report, drafted the project report and edited the manuscript.
PB was responsible for data collection and data entry, conducted the data analyses, contributed to the project report and edited the manuscript.
JH participated in the design of the study, the supervision of the data collection, interpretation of the results, production of the project report and editing the manuscript.
PK oversaw the administration of the grant and edited the manuscript.
HS was Convenor and Chair of the Steering Committee, and participated in editing the manuscript.
LF participated in the design of the study, the supervision of the data collection, interpretation of the results, production of the project report and the drafting of the manuscript
Acknowledgements
This project was funded by the National Cancer Control Initiative (NCCI) and we would particularly like to thank Mark Elwood for his support. We are grateful to the members of the WA Clinical Oncology Group and the NCCI as well as other Western Australian pathologists and clinicians who contributed to this report. The cooperation and input of the staff at the WACR and HBCRs to this project are gratefully acknowledged.
==== Refs
Sharma T Kumar A Manocha KK Staging of malignant diseases and its significance Haryana Medical Journal 2001 21
Gospodarowicz MK Henson DE Hutter RVP O'Sullivan B Sobin LH Wittekind Ch eds Prognostic Factors in Cancer 2001 Wiley-Liss, USA
Threlfall TJ Brameld K Cancer survival in Western Australian Residents, 1982–1997 2000 Health Department of Western Australia, Perth Statistical series number 60
American Joint Committee on Cancer Cancer Staging Manual 2002 Sixth Springer-Verlag, New York
Gospodarowicz MK Benedet L Hutter RV Fleming I Henson DE Sobin LH History and international developments in cancer staging Cancer Prevention and Control 1998 2 262 268 10470455
O'Sullivan B Gospodarowicz MK Mackillop WJ Evans WK Whylie B Ashbury FD The consultation to develop a national strategy for cancer staging in Canada Cancer Prevention and Control 1998 2 287 294 10470458
Threlfall TJ Thompson JR Cancer incidence and mortality in Western Australia, 2002 2004 Department of Health, Western Australia, Perth Statistical series number 71
Esteban D Whelan S Laudico A Parkin DM Manual for Cancer Registry Personnel. International Agency for Research on Cancer (IARC), World Health Organization (WHO) and International Association of Cancer Registries (IACR) Lyon 1995 IARC Technical Report no. 10
Clark WH The developmental biology of primary cutaneous malignant melanoma Seminars in Oncology 1975 2 83 790575
Breslow A Thickness, cross-sectional area and depth of invasion in the prognosis of cutaneous melanoma Annals of Surgery 1970 172 902 908 5477666
Boutard P Platell C Threlfall T Model for collecting colorectal cancer staging information in Western Australia ANZ J Surg 2004 74 895 899 15456441 10.1111/j.1445-1433.2004.03199.x
|
16105180
|
PMC1232866
|
CC BY
|
2021-01-04 16:37:42
|
no
|
Popul Health Metr. 2005 Aug 17; 3:9
|
utf-8
|
Popul Health Metr
| 2,005 |
10.1186/1478-7954-3-9
|
oa_comm
|
==== Front
Theor Biol Med ModelTheoretical Biology & Medical Modelling1742-4682BioMed Central London 1742-4682-2-351613732410.1186/1742-4682-2-35ResearchNucleic acid chaperons: a theory of an RNA-assisted protein folding Biro Jan C [email protected] Homulus Foundation, 88 Howard, #1205; San Francisco, 94 105 CA, USA2005 1 9 2005 2 35 35 10 7 2005 1 9 2005 Copyright © 2005 Biro; licensee BioMed Central Ltd.2005Biro; 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
Proteins are assumed to contain all the information necessary for unambiguous folding (Anfinsen's principle). However, ab initio structure prediction is often not successful because the amino acid sequence itself is not sufficient to guide between endless folding possibilities. It seems to be a logical to try to find the "missing" information in nucleic acids, in the redundant codon base.
Results
mRNA energy dot plots and protein residue contact maps were found to be rather similar. The structure of mRNA is also conserved if the protein structure is conserved, even if the sequence similarity is low. These observations led me to suppose that some similarity might exist between nucleic acid and protein folding. I found that amino acid pairs, which are co-located in the protein structure, are preferentially coded by complementary codons. This codon complementarity is not perfect; it is suboptimal where the 1st and 3rd codon residues are complementary to each other in reverse orientation, while the 2nd codon letters may be, but are not necessarily, complementary.
Conclusion
Partial complementary coding of co-locating amino acids in protein structures suggests that mRNA assists in protein folding and functions not only as a template but even as a chaperon during translation. This function explains the role of wobble bases and answers the mystery of why we have a redundant codon base.
==== Body
Introduction
The protein folding problem has been one of the grand challenges in computational molecular biology. The problem is to predict the native three-dimensional structure of a protein from its amino acid sequence. It is widely believed that the amino acid sequence contains all the necessary information for the correct three-dimensional structure, since protein folding is apparently thermodynamically determined; i.e., given a proper environment, a protein will fold spontaneously to the correct conformation. This is called Anfinsen's thermodynamic principle [1].
The thermodynamic principle has been confirmed many times on many different kinds of proteins in vitro. Critics says that the in vivo chemical conditions are different from those in vitro, correct protein folding is determined by interactions with other molecules (chaperons, hormones, substrate, etc.) and is much more complex than renaturation of denatured poly amino acids. The fact that many naturally-occurring proteins fold reliably and quickly to their native states, despite the astronomical number of possible configurations, has come to be known as Levinthal's paradox [2].
Anfinsen's principle was formulated in the 1960s using purely chemical experiments and a lot of intuition. Today, many sequences and structures are available to establish a logical and understandable link between sequence, structure and function. But it is still not possible to predict the structure (or a range of possible structures) correctly from the sequence alone, ab initio and in silico [3].
There are two potential, external sources of additional and specific protein folding information: (a) the chaperons (other proteins that assist in the folding of proteins and nucleic acids [4]; and (b) the protein-coding nucleic acid sequences themselves (which are templates for protein syntheses, but are not defined as chaperons). Protein chaperons are not necessarily similar to their clients; they can be complementary templates, too, as it is well known from nucleic acid interactions. However, chaperons necessarily contain spatial information (in some form) that guides another protein to fold correctly. Chaperoning requires subtle interactions with the immaturely folded intermediate so that its structure is loosened and it is then released for successive rounds of folding attempts. (Some aspects of this situation might be compared to enzyme-substrate interactions and kinetics.)
The possibility that the nucleotide sequence itself could modulate translation and hence affect co-translational folding and assembly of proteins has been investigated in a number of studies [5-7]. Studies on the relationships between synonymous codon usage and protein secondary structural units are especially popular [8-10]. The genetic code is redundant (61 codons encode 20 amino acids) and as many as six synonymous codons can encode the same amino acid (Arg, Leu, Ser). The "wobble" base has no effect on the meaning of most codons, but codon usage (wobble usage) is nevertheless not randomly defined [11,12] and there are well-known, stable species-specific differences in codon usage. It seems to be reasonable to search for the meaning (biological purpose) of the wobble bases in association with protein folding.
Materials and methods
We have developed a tool, SeqX [13], which is specially designed to provide 2D projections of protein structures (residue contact maps) and analyze residue co-locations statistically in these structures. We have collected residue co-location statistics (residue contact statistics) from 80 different structures from the Protein Data Bank (PDB) [14]. This non-redundant SeqX data set listed ~35,000 amino acid co-locations (i.e. residues located within a 6 Å radius of the alpha carbon atoms; neighbor residues on the same strand were excluded).
The mfold tool was used to obtain RNA structure data [15] and the energy dot-plots provided by this program were used to estimate the site and size of the most probable RNA folding.
Student's t-tests were used for statistical evaluation.
Results and Discussion
The very first idea of protein folding on a nucleic acid template was the model of direct protein synthesis on the surface of dsDNA. It was suggested by George Gamow [16,17] before the discovery of the genetic code and mRNA. Gamow noticed that the distances between base pairs in DNA and the distances between amino acids in proteins are the same (~4 Å), and he suggested that complementary base pairs formed 20 different "cavities" on the surface of the DNA (one for each amino acid) where the amino acid residues aligned and became ligated. The correct translation turned out to be mRNA-mediated and stereochemical fitting between DNA and protein residues was rejected [18].
However, this question arose again in a different form. Specific DNA-protein interactions do exist (such as those between DNA and transcription factors or between restriction enzymes and recognition sequences), and it is difficult to explain the extreme specificity of these interactions without assuming that there is a "small scale", "residue level" interaction between nucleic acids and proteins. I found Woese's idea [19] of stereochemical fitting very attractive, i.e. affinity between codons and coded amino acids, in contrast to Crick's statement of a "frozen accident" [18]. I succeeded in constructing A Common Periodic Table of Codons and Amino Acids [20] and in showing a large number of codon-amino acid co-locations in restriction enzyme-recognition sequence structures [21]. Consequently, I support the view that the unit of specific nucleic acid-protein interactions is the codon and its amino acid.
Nucleic acids are structure-forming molecules. Perfect complementarity between Watson and Crick (WC) base pairs forms the perfect helical structure, dsDNA. However, partial or suboptimal WC complementarity in and between strands provides a large number of DNA/RNA structure variations. The structural variation of a given RNA might be large; some structures are energetically more favored, some are less. The importance of one RNA secondary structure over another is usually not a subject of debate, because RNA structure often has no known physiological significance (there are exceptions, e.g. tRNAs).
Proteins are also structure-forming molecules. However, in contrast to nucleic acids, there is no known specific amino acid complementarity, and the known physicochemical rules (charge, hydrophobe, size compatibility) are often insufficient to define only one obvious protein folding and structure (Biro, 2005, unpublished). The limitation of Anfinsen's theorem [1] is described by the Levinthal paradox [2], which is confirmed by the often frustrating outcome of ab initio protein prediction. However, we know that there is very little biological tolerance for variation in protein structure; usually only one main functioning structure is assigned to a protein sequence (and sometimes a few allosteric variants). The exact structure of a protein is critical, as is evident from our knowledge of prions. However, the primary sequence is usually insufficient to establish this exact structure and chaperons are required. The problem is not that there is a large choice of different protein folding pathways with different end-points, only one of which is physiologically normal. Rather, the problem is the risk of deviation from the (physiological) folding pathway to form any one of a number of misfolded molecules. Chaperones are needed because the sequence is insufficient to define the most effective folding pathway leading to the thermodynamically most stable structure.
Chaperons are defined as proteins of which the function is to assist the folding of other proteins. However, the most obvious chaperons for me are nucleic acids; specifically, those coding the protein in question (Fig. 1). Immediate RNA-assisted protein folding prevents any protein misfolding at the site of protein synthesis itself. The insufficiency of folding information in protein sequences is more than compensated by the excess of information (codon base redundancy) in nucleic acids.
Figure 1 Nucleo-protein structures and protein foldings. The distance between bases in nucleic acids (horizontal blue lines) and amino acids in proteins (red dots) is almost the same (A). This suggests the possibility of residue-level interactions between these molecules (B). Partial, sub-optimal complementarity between DNA strands ("honeycomb structure") and fitting of amino acids into DNA cavities was suggested by Gamow [16, 17] (C, D). A further development of this model is that partial complementarity between mRNA subsequences (E, F) determines the orientation of amino acid residues in ribonucleoprotein complexes and consequently the RNA loops serve as templates (RNA chaperons) to main secondary protein structures such as alpha helices (G) and beta-sheets (H). Codon boundaries are not indicated in these models. The Figure illustrates the historical development of the concept of direct and specific nucleic acid – protein interactions and its possible consequences for protein folding.
I compared the structures of mRNAs with those of the translated proteins to test the assumption that protein folding information is present in mRNA. The energy dot-plots provided by mfold and the 2D protein structures provided by SeqX indeed suggest similarity in most of the randomly selected structures (Fig. 2)
Figure 2 Comparison of 12 randomly selected protein and corresponding mRNA structures. Residue contact maps (RCM) were obtained from the PBD files of the protein structures using the SeqX tool (left triangles). Energy dot plots (EDP) for the coding sequences were obtained using the mfold tool (right triangles). The two maps were aligned along a common left diagonal axis to facilitate visual comparison between the different possible representations. The black dots in the RCMs indicate amino acids that are within 6 Å of each other in the protein structure. The colored (grass-like) areas in the EDPs indicate the energetically mostly likely RNA interactions (color code in increasing order: yellow, green red, black). The full names and the lengths of the proteins (number of amino acid residues): 1AM5: PEPSIN (324), 1A8D: TETANUS NEUROTOXIN (451), 1MD8: SERIN PROTEASE (329), 1ARB: ACHROMOBACTER PROTEASE I (268), 1HO9: A ALPHA-2A ADRENERGIC RECEPTOR (32), 1BIA: BIRA BIFUNCTIONAL PROTEIN (376), 1CWN: ALDEHYDE REDUCTASE (324), 1BG4: ENDO-1,4-BETA-XYLANASE (302), 1SIG: RNA POLYMERASE PRIMARY SIGMA FACTOR (339) bases, 1K40: ADHESION KINASE (126), 1EZJ: NUCLEOCAPSID PHOSPHOPROTEIN (140), 1ABN: ALDOSE REDUCTASE (315). The coordinates indicate the number of amino acid and the corresponding nucleic acid residues.
Another similar, but still not quantitative, comparison of protein and coding structures was performed on four proteins that are known to have very similar 3D structures although their primary structures (sequences) are less than 30% similar, and on the sequences of their mRNAs. These four proteins exemplify the fact that protein tertiary structure is much more conserved than the amino acid sequence. I asked whether this is also true for RNA structures and sequences. I found that there are signs of conservation even of RNA secondary structure (as indicated by the energy dot plots) and there are similarities between the protein and nucleic acid structures (Figure 3).
Figure 3 Comparison of the protein and mRNA secondary structures. Residue contact maps (RCM) were obtained from the PBD files of four protein structures (1CBI, 1EIO, 1IFC, 1OPA) using the SeqX tool (left column). Energy dot plots (EDP) for the coding sequences were obtained using the mfold tool (right column). The left diagonal portions of these two maps are compared in the central part of the figure. Blue horizontal lines in the background correspond to the main amino acid co-location sites in the RCM. Intact RNA (123) as well as subsequences containing only the 1st and 3rd codon letters (13) are compared. The black dots in the RCMs indicate amino acids that are within 6 Å of each other in the protein structure. The colored (grass-like) areas in the EDPs indicate the energetically most likely RNA interactions (color code in increasing order: yellow, green red, black). The full names and the lengths of the proteins (number of amino acid residues): 1CBI: CELLULAR RETINOIC ACID BINDING PROTEIN I (136), 1EIO: ILEAL LIPID BINDING PROTEIN (127), 1IFC: INTESTINAL FATTY ACID BINDING PROTEIN (132), 1OPA: CELLULAR RETINOL BINDING PROTEIN II (135).
These structural comparisons are suggestive, but not quantitative, and more convincing statistical evaluation is necessary to evaluate the significance of the suggested similarity between nucleic acid and corresponding protein structures. (Quantitative comparisons of 2D protein representations and RNA energy dot plots are possible and are in progress in our laboratory). Similarities between two macromolecules (RNA-RNA, protein-protein, RNA-protein), or even between two macromolecular families, does not automatically mean that they are functionally related to each other (or that one is a chaperon), but it is a widely accepted sign of a biologically significant relationship.
The molecular basis of mRNA structure formation is the known WC base pair complementarity. Therefore I asked whether it is possible to find some kind of complementarity between the codons of co-locating (specifically interacting) amino acids.
Searching for some pattern in the codons of co-locating amino acids, the frequency of the eight possible patterns in the 64 nucleic acid triplets was analyzed. The codons were either complementary to each other in all three (-123-) or in at least two codon base positions (-12X-, -1X3-, -X23-). In these latter cases the codon complementarity was partial, because complementarity was not required for one position (X). The complementary codons were translated in the same (5' > 3' and 3' > 5', only complementary, C) or the reversed and complementary (5' > 3' and 5' > 3', RC) directions. One (and only one) codon complementary pattern of the eight possible turned out to be significantly overrepresented among the codons of co-locating amino acids: D-1X3/RC-3X1. The other 7 possible codon patterns served as negative controls. This pattern means that the 1st and 3rd codon residues are complementary in reverse orientation, but the 2nd residue may be but is not necessarily complementary (X) (Fig. 4). The possible amino acid pairs determined by the D-1X3/RC-3X1 formula are indicated in Table I.
Figure 4 Complementary codes vs. amino acid co-locations. (A) The propensity of the 400 possible amino acid pairs was monitored in 80 different protein structures with the SeqX tool. The tool detected co-locations when two amino acids were closer than 6 Å to each other (neighbors on the same strand were excluded). The total number of co-locations was 34,630. Eight different complementary codes were constructed for the codons (two optimal and six suboptimal). In the two optimal codes all three codon residues (123) were complementary (C) or reverse-complementary (RC) to each other. In the suboptimal codes only two of three codon residues were C or RC to each other (12, 13, 23), while the third was not necessarily complementary (X). (For example, complementary code RC_1X3 means that the first and third codon letters are always complementary, but not the second, and the possible codons are read in reverse orientation). The 400 co-locations were divided into 20 subgroups corresponding to 20 amino acids (one of the co-locating pairs), each group containing the 20 amino acids (corresponding to the other amino acid in the co-locating pair). If the codons of the amino acid pairs followed the predefined complementary code, the co-location was regarded as positive (P); if not, the co-location was regarded as negative (N). Each symbol represents the mean frequency of P or N co-locations corresponding to the indicated amino acid. Paired Student's t-test, n = 20 (see Fig. 2 for explanation). (B) The ratio of positive (P) and negative (N) co-locations was calculated on data from (A). Each bar represents the mean ± SEM, n = 20.
Table I Amino Acids Coded by Partially Complementary Codons
D_1X3/RC_3X1 1st G T G G T G C A A CT A A C C AC AT A G T T
2nd C G A A T G A T A T T A C A G CG C T G A
3rd X CT CT AG CT X CT ACT AG XAG G CT X AG AGX CTX X X G CT
1st 2nd 3rd AA A C D E F G H I K L M N P Q R S T V W Y
G C X A + + + + + + + + + + + + + + +
T G CT C + + + + + + + +
G A CT D + + + + + + + +
G A AG E + + + + + + + +
T T CT F + + + + + + + +
G G X G + + + + + + + + + + + + + + +
C A CT H + + + + + + + +
A T ACT I + + + + + + + + + + +
A A AG K + + + + + + + +
CT T XAG L + + + + + + + + + + + + +
A T G M + + + +
A A CT N + + + + + + + +
C C X P + + + + + + + + + + + + +
C A AG Q + + + + + +
AC G AGX R + + + + + + + + + + + + + + + + +
AT CG CTX S + + + + + + + + + + + + + +
A C X T + + + + + + + + + + + + + + +
G T X V + + + + + + + + + + + + + + +
T G G W + + + +
T A CT Y + + + + + + + +
RC_3X1 code: 1st and 3rd codon letters are complementary in reverse order, indicated by complementary colors (red, blue); X: any residue; AA: amino acids, one-letter code, +: AAs coded by the D_1x3/RC_3X1 complementary codons
This partial, suboptimal complementarity again suggests that mRNA folding may assist protein folding, but does not necessarily prove it. An alternative explanation is that it is only a sign of the biochemical origin of specifically interacting amino acid pairs (they are encoded in partially complementary codons) but does not mean that complementary structures in amino acids will form interacting protein strands.
The historical concept of specific nucleic acid – protein interactions and the subsequent possibility of RNA-assisted protein folding was illustrated in figure 1. I wish to suggest a further development of these ideas. The distance between codons is about three times larger than the distance between amino acids and therefore complete 1 by 1 RNA-protein alignment is not possible. Furthermore, a long continuous alignment would create problems in dissociating the nucleoprotein complexes. Therefore I suggest that only some basic (positively charged) amino acids remain attached to their codons (or become re-attached after removal of tRNA). If this attachment point is followed by a loop in the mRNA, a corresponding loop will be formed in the nascent protein (Figure 5). The interaction between the positively charged amino acid and the negatively charged codon will be successively weakened by the growing protein loop and finally interrupted, for example, by the translation of a negatively charged amino acid. It is known that interactions between nucleic acids and proteins often involve only a few amino acids and that these "patchy" interaction sites often contain an arginine [21]. Complex protein structures might be folded in this way (Figure 6).
Figure 5 RNA assisted protein loop formation. Translation begins with the attachment of the 5' end of a mRNA to the ribosome (A). Ribonucleotides are indicated by blue + and the 1st and 3rd bases in the codons by blue lines, while the 2nd base positions are left empty. A positively charged amino acid [(+) and red dots], for example arginine, remains attached to its codon. The mRNA forms a loop because the 1st and 3rd bases are locally complementary to each other in reverse orientation (B). The growing protein is indicated by red circles (o). When translation proceeds to an amino acid with especially high affinity to the mRNA-attached arginine, for example a negatively charged Glu or Asp [(-) and blue dot], the charge attraction removes the Arg from its mRNA binding site and the entire protein is released from the mRNA and completes a protein loop (C). The protein continues to grow toward the direction of its carboxy terminal (COOH).
Figure 6 RNA-assisted (translational) protein folding. There are three reverse and complementary regions in a mRNA (blue line, A): a-a', b-b', c-c', which fold the mRNA into a T-like shape. During the translation process the mRNA unfolds on the surface of the ribosome, but subsequently refolds, accompanied by its translated and lengthening peptide (red dotted line, B-F). The result of translation is a temporary ribonucleotide complex, which dissociates into two T-shape-like structures: the original mRNA and the properly folded protein product (G). The red circles indicate the specific, temporary attachment points between the RNA and protein (for example a basic amino acid) while the blue circles indicate amino acids with exceptionally high affinity for the attachment points (for example acidic amino acids); these capture the amino acids at the attachment point and dissociate the ribonucleoprotein complex. Transfer-RNAs are of course important participants in translation, but they are not included in this scenario.
The observed partial complementary coding of co-locating amino acids (the D_1X3/RC_3X1 formula) raises a series of interesting questions. The 20 amino acid – triplet codon model, obviously entails the need for a third codon base (two nucleotides are simply not enough). However, based on the assumption of RNA chaperons, two proteins with identical primary structures (for example human and chimpanzee Hb) may fold differently if there are differences in the redundant codon base positions. Similarly, a number of SNPs (Single Nucleotide Polymorphisms) that do not change the coded amino acids may result in protein structure variations.
The medical genetics literature (for example OMIM) is full of annotations concerning wobble base mutations and it is usually inferred that these "translationally silent" mutations are unlikely to cause disease. A famous exception is prion diseases (mad cow disease, Creutzfeldt-Jakob disease [22]). This large group of diseases is characterized by the presences of an abnormally folded protein (PrPsc) instead of the normally folded one (PrPC). The physiological and abnormal proteins have the same primary structures; only the secondary structures are different. In most cases the disease is acquired by infection, but there are many inherited forms. At least 42 known point mutations, 24 causative and 18 translationally silent, are described in the literature [23]. The wobble base mutations demand serious attention, especially since it is known that selection pressure exists for the wobble bases in some codon positions [24].
The RNA chaperon theory does not mean that every wobble-base point-mutation (or SNP) influences secondary structure. Usually, many codons and amino acids are involved in the formation of a simple secondary structure element (helix, sheet, turn) and probably most mutations have no structural consequences. Also, many mutations are accompanied by a second, compensatory mutation that corrects the structural consequences of the first. In evolution, sequence changes more rapidly than structure; however, many sequence changes are compensatory and preserve local physicochemical characteristics. For example, if an amino acid side chain is particularly bulky with respect to the average at a given position in a given sequence, this might have been compensated in evolution by a particularly small side chain in a neighbouring position, preserving the general structural motif [25].
An additional question raised by the RNA-chapeon hypotheses concerns the GC versus AT contents of various genomes, which range from 78 / 22 to 22 / 76. This causes marked differences, especially in the compositions of the third codon nucleotides. It is reasonable to suppose that redundant codon bases are susceptible to much more variation if there is no amino acid replacement, and that if such changes affect protein folding, this would have restrained such nucleotide replacements significantly. However this is not necessarily true. The partial complementary coding of co-locating amino acids (the D_1X3/RC_3X1 rule) suggests that the number of possible amino acid co-locations is less than 200 (20 × 20/2), and the possible co-locations involve pairings of physicochemically compatible amino acids (Biro, 2005, unpublished). Many non-silent mutations in one codon are coupled to a second (silent or non-silent) mutation in a second codon. This coupled and coordinated model of mutations actually permits a very large number of variations in the primary nucleic acid and protein sequences with no consequences for nucleic acid or protein secondary structures. And as indicated above, 3D structures are generally much more conserved than sequences.
Complementary coding of co-locating amino acids, and the consequent possibility of nucleic acid assisted protein folding (nucleic acid chaperon), might give new insights into the dilemma of why we have a redundant codon base and might explain the role of the wobble base in the codon. Experimental, in vitro support is necessary to confirm this in silico suggestion of nucleic acid chaperons.
Declaration of competing interests
The author(s) declare that they have no competing interests.
==== Refs
Anfinsen CB Redfield RR Choate WI Page J Carroll WR Studies on the gross structure, cross-linkages, and terminal sequences in ribonuclease J Biol Chem 1954 207 201 210 13152095
Levinthal C Debrunner P, Tsibris JCM, Munck E How to fold graciously in Mossbauer spectroscopy in biological systems Proceedings of a Meeting held at Allerton House, Monticello, IL 1969 Urbana, IL: University of Illinois Press 22 24
Klepeis JL Floudas AC ASTRA-FOLD: a combinatorial and global optimization framework for ab initio prediction of three-dimensional structures of proteins from the amino acid sequence Biochem J 2003 85 2119 2146
Walter S Buchner J Molecular chaperones – cellular machines for protein folding Angew Chem Int Ed Engl 2002 41 1098 1113 12491239 10.1002/1521-3773(20020402)41:7<1098::AID-ANIE1098>3.0.CO;2-9
Komar AA Kommer A Krasheninnikov IA Spirin AS Cotranslational folding of globin J Biol Chem 1997 272 10646 10651 9099713 10.1074/jbc.272.16.10646
Thanaraj TA Argos P Protein secondary structural types are differentially coded on messenger RNA Protein Sci 1996 5 1973 1983 8897597
Brunak S Engelbrecht J Protein structure and the sequential structure of mRNA: alpha-helix and beta-sheet signals at the nucleotide level Proteins 1996 25 237 252 8811739 10.1002/(SICI)1097-0134(199606)25:2<237::AID-PROT9>3.3.CO;2-Y
Gupta SK Majumdar S Bhattacharya TK Ghosh TC Studies on the relationships between the synonymous codon usage and protein secondary structural units Biochem Biophys Res Commun 2000 269 692 696 10720478 10.1006/bbrc.2000.2351
Chiusano ML Alvarez-Valin F Di Giulio M D'Onofrio G Ammirato G Colonna G Bernardi G Second codon positions of genes and the secondary structures of proteins. Relationships and implications for the origin of the genetic code Gene 2000 261 63 69 11164038 10.1016/S0378-1119(00)00521-7
Gu W Zhou T Ma J Sun X Lu Z The relationship between synonymous codon usage and protein structure in Escherichia coli and Homo sapiens Biosystems 2004 73 89 97 15013221 10.1016/j.biosystems.2003.10.001
Ermolaeva O Synonymous codon usage in bacteria Curr Issues Mol Biol 2001 3 91 97 11719972
Biro JC Biro JM Biro AM Hidden massages in hidden sub-sequences: a study on collagens 30th FEBS Congress – 9th IUBMB Conference, Budapest, Hungary, 2–7 July 2005 2005 abstract.
Biro JC Fördös G SeqX: a tool to detect, analyze and visualize residue co-locations in protein and nucleic acid structures BMC Bioinformatics 2005 [SeqX: ].
Berman HM Westbrook J Feng Z Gilliland G Bhat TN Weissig H Shindyalov IN Bourne PE The protein data bank Nucleic Acids Res 2000 28 235 242 [PDB ]. 10592235 10.1093/nar/28.1.235
Zuker M Mfold web server for nucleic acid folding and hybridization prediction Nucleic Acids Res 2003 31 3406 3415 12824337 10.1093/nar/gkg595
Gamow G Possible relation between deoxyribonucleic acid and protein structures Nature 1954 173 318
Gamow G Possible mathematical relation between deoxyribonucleic acid and proteins Biol Med 1954 22 1 13
Crick FHC The origin of the genetic code J Mol Biol 1968 38 367 379 4887876 10.1016/0022-2836(68)90392-6
Woese CR The Genetic Code: The Molecular Basis for Gene Expression 1967 New York: Harper & Row 156 160
Biro JC Benyo B Sansom C Szlavecz A Fordos G Micsik T Benyo Z A common periodic table of codons and amino acids Biochem Biophys Res Commun 2003 306 408 415 12804578 10.1016/S0006-291X(03)00974-4
Biro JC Biro JM Frequent occurrence of recognition site-like sequences in the restriction endonucleases BMC Bioinformatics 2004 5 30 15113406 10.1186/1471-2105-5-30
Prusiner SB Prions PNAS 1998 95 13363 13383 9811807 10.1073/pnas.95.23.13363
The Official Mad Cow Disease Home Page – Known point variations in human prion gene coding region – Last updated 16 Nov 2000
Luck R Steger G Riesner D Secondary structure of prion mRNA J Mol Biol 1996 258 813 26 8637012 10.1006/jmbi.1996.0289
Neher E How frequent are correlated changes in families of protein sequences? Proc Natl Acad Sci USA 1994 91 98 102 8278414
|
16137324
|
PMC1232867
|
CC BY
|
2021-01-04 16:39:26
|
no
|
Theor Biol Med Model. 2005 Sep 1; 2:35
|
utf-8
|
Theor Biol Med Model
| 2,005 |
10.1186/1742-4682-2-35
|
oa_comm
|
==== Front
Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-651610721710.1186/1743-422X-2-65ResearchEvidence of HIV exposure and transient seroreactivity in archived HIV-negative severe hemophiliac sera Tenenbaum Scott A [email protected] Cindy A [email protected] Steve S [email protected] Harris E [email protected] Robert F [email protected] Cindy A [email protected] Department of Biomedical Sciences, Ge*NY*Sis Center for Excellence in Cancer Genomics, University at Albany-SUNY, Albany, NY, USA2 Microbiology and Immunology, Tulane University School of Medicine, Tulane University School of Medicine, New Orleans, LA, USA3 Ortho Diagnostic Systems, HlV and Hepatitis Research and Development, Raritan, NJ, USA4 Department of Medicine, Section of Hematology and Medical Oncology, Tulane University School of Medicine, New Orleans, LA, USA2005 17 8 2005 2 65 65 8 8 2005 17 8 2005 Copyright © 2005 Tenenbaum et al; licensee BioMed Central Ltd.2005Tenenbaum 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
Approximately 25% of hemophiliacs that were frequently exposed to blood clotting factor concentrates (CFCs) contaminated with human immunodeficiency virus (HIV) are presently HIV seronegative. In this study, we sought to determine if some of these individuals were at any time transiently HIV seropositive. In the early to mid-1980s the majority of severe hemophilia patients were exposed to CFCs contaminated with HIV. Although many of these hemophiliacs became HIV-positive, a small percentage did not become infected. To determine if some of these individuals successfully resisted viral infection, we attempted to document the presence of transient HIV reactive antibodies in archived plasma samples (1980–1992) from currently HIV-negative severe hemophiliacs who had a high probability of repeated exposure to HIV contaminated CFC. Archived plasma samples were retrospectively tested using an FDA approved HIV-1Ab HIV-1/HIV-2 (rDNA) enzyme immunoassay (EIA) and a HIV-1 Western blot assay (Wb), neither of which were commercially available until the late 1980s, which was after many of these samples had been drawn.
Results
We found that during the high risk years of exposure to HIV contaminated CFC (1980–1987), low levels of plasma antibodies reactive with HIV proteins were detectable in 87% (13/15) of the haemophiliacs tested. None of these individuals are presently positive for HIV proviral DNA as assessed by polymerase chain reaction (PCR).
Conclusion
Our data suggest that some severe hemophiliacs with heavy exposure to infectious HIV contaminated CFC had only transient low-level humoral immune responses reactive with HIV antigens yet remained HIV-negative and apparently uninfected. Our data supports the possibility of HIV exposure without sustained infection and the existence of HIV-natural resistance in some individuals.
==== Body
Background
In the 1980's an estimated 17,000 people in the United States were affected the congenital blood clotting factor deficiencies, Hemophilia A and B (Factor VIII and Factor IX deficiency, respectively). Since the early 1970's, the mainstay of treatment for bleeding in hemophilia patients has been the use of clotting factor concentrates (CFCs) commercially prepared from large plasma pools comprised of thousands of individual donors. Prior to 1985 CFCs were prepared from donors with unknown HIV infection status and were not routinely subjected to viral inactivation procedures. With each infusion from a new lot of clotting factor concentrate, hemophilia patients were exposed to plasma from approximately 2,000 to 25,000 donors [1]. As a result, roughly 50% of the total hemophilia population in the United States became infected with HIV prior to the institution of donor screening and the use of viral inactivation procedures of factor concentrates in 1985 [2-4]. Since 1987 there has been a virtual elimination of HIV-1 infection in the hemophilia population [3-6].
Largely due to the extensive network of comprehensive hemophilia treatment centers, the hemophilia population has been actively studied for possible variables that may influence HIV infection and progression. Retrospective analysis of hemophiliac plasma samples stored as part of routine clinical visits has shown that HIV infection, as documented by permanent HIV-seroconversions began in 1978, peaked in 1982, and ended by 1987. In general, those patients who received the greatest exposure to CFCs were at the highest risk for HIV infection [7]. Hemophiliacs exposed to factor-VIII concentrates, in general, were more likely to become infected than those exposed to factor-IX concentrates (prothrombin complex concentrates or PCCs). Patients who received an average of over 20,000 units of factor-VIII concentrate annually during the early 1980's had a cumulative incidence of HIV-infection exceeding 90% and those receiving comparable doses of PCCs had a cumulative incidence exceeding 50% [3,4]. This clearly demonstrates the prevalence of infectious HIV in the United States CFC supply.
Not all hemophiliacs exposed to CFCs contaminated with infectious HIV were ultimately infected with the HIV virus. Although inoculum size may account for the lack of infection in some hemophiliacs, factors such as age, race, sex or pre-existing medical condition has not been found to be related to risk of HIV infection. However, several studies have shown that certain HLA types were associated with either an increased or decreased risk of HIV infection in hemophilia patients [3,8-10].
In 1996, three independent groups identified the chemokine co-receptor 5 (CCR-5) as a secondary receptor for the HIV virus. The presence of two copies of a naturally occurring deletion mutation of the CCR-5 receptor (CCR-5Δ32) apparently conferred resistance to infection by the virus [11-14]. Heterozygous expression of CCR-5Δ32 did not appear to prevent HIV-1 infection but may have resulted in slower decline in CD4+ cells, lower levels of plasma viremia and in slower progression to AIDS [15-18]. From 1979–1985 severe hemophiliacs, as defined as individuals producing less than 1% of the normal value of a clotting factor, were exposed to the largest volume of clotting factor concentrates. Accordingly, 90% of these individuals became infected with HIV [4]. However, some severe hemophiliacs have remained H [V-negative despite repeated exposure to CFCs. We hypothesized that information concerning natural HIV resistance might be obtained by the investigation of such high-risk individuals. Although HIV enzyme linked immunoassays (EIA) and Western blot assays (Wb) were available for detecting antibodies reactive with HIV antigens as early as 1984, the specificities and sensitivities of these immunoassays have increased dramatically in more contemporary versions of these diagnostic tests [19-23]. We speculated that some High-risk hemophiliacs exposed to contaminated CFCs, although not permanently infected, may have mounted transient humoral and/or cellular immune responses reactive with HIV proteins but were below the threshold of delectability of the earlier first generation HIV diagnostic tests. Using more sensitive HIV immunologic diagnostic tests, we reassessed anti-HIV reactivity in archived plasma samples (1980–1992) from presently HIV-seronegative severe hemophiliacs that had exposure to large quantities of contaminated CFCs (HIV exposed/-HIV negative hemophiliacs).
Results
Detection of transient HIV reactive plasma antibodies
We had adequate yearly representation and sample quantity in our archived collection to assemble plasma sets from 15 severe hemophiliacs with extremely likely exposure to HIV contaminated clotting concentrates (Table 1). All of the archived plasma sets tested contained samples that were collected prior to 1986. Using improved HIV-EIA or HIV-Wb immunoassay analysis we detected antibodies reactive with HIV antigens in one or more samples from 13/15 (87%) of the archived plasma sets tested (Figure 1, panels 2-4 and Table 2, patients 1–13). Of these, two plasma sets had samples reactive to both the HIV-EIA and the HIV-WB (Figure 1, panel 2 and Table 2, patient 2 and 5). An additional three plasma sets contained samples displaying reactivity only on the HIV-EIA (Table 2, patients 1, 6 and 7) which occurred in samples collected in 1987 or earlier. Eight of the hemophilia patient archived plasma sets contained samples with antibodies (IgG and/or IgM) only reactive with one or more HIV antigen as determined by HIV-1 Wb analysis (representative examples panels 2–4, Figure 1 and Table 2, patients 3, 4, 8–13). HIV proteins were recognized by antibodies in HIV-1 Wb reactive plasma samples in the following frequencies; group antigen (Gag) p17 (3/10, 30%), group antigen (Gag) p24 (8/10, 80%), reverse transcriptase (RT) p66 (3/10, 30%), envelope glycoproteins (Env) gp120 and 160 (2/10, 20%), One archived plasma set contained a sample that displayed reactivity to multiple HIV antigens on HIV-1 Wb analysis that approached meeting the criteria for "HIV-positive" reactivity (Figure 1, panel 4 and table 2, patient 4). The observed multiple HIV protein reactivity was only present in the 1983 sample from this archived plasma set. As a whole, indeterminate banding activity on the HIV-1 Wb appeared to fluctuate in intensity within many of the archived plasma sets that displayed reactivity (Figure 1 panels 1–3 and Table 2 asterisks). The greatest intensity of banding activity in these archived plasma sets coincided with years of highest-risk of exposure to HIV contaminated clotting factor concentrate (1980–1985). None of the HIV-1 Wb indeterminate hemophiliacs showed reactivity to HIV-2 proteins when analyzed by HIV-2 Wb analysis (data not shown).
Figure 1 HIV-1 Western blot reactivity of HIV-exposed/HIV seronegative hemophiliacs. Panels 1–4: HIV-1 western blot analysis of chronologically archived plasma samples from HIV-exposed/HIV-seronegative hemophiliacs displaying transient partial reactivity against HIV-1 proteins. Panel 5: Chronological serum samples from an HIV-1 seropositive hemophiliac. Panel numbers (1–4) correspond to 1–4 respectively in tables 1 and 2.
Table 1 Clinical profile and clotting factor concentrate exposure of HIV-Exposed/HIVseronegative hemophiliacs.
Hemophilia Factor_concentrate exposure (1980–1985)
Patient # Year of birth Type Severity* Presence of inhibitor** Type Average Units/year
1 1972 B <1% No PCC 55,000
2 1956 B 4% No PCC <10,000
3 1966 A <1% Yes PCC 27,000
4 1953 A <1% No VIII 15,000
5 1977 B <1% No PCC 29,000
6 1971 A <1% Yes PCC 30,000
7 1931 A <1% No VIII 11,000
8 1976 B <1% No PCC 18,000
9 1964 B <1% No PCC 24,000
10 1976 A <1% Yes PCC 12,000
11 1957 A <1% No VIII 29,000
12 1950 A <1% Yes PCC 80,000
13 1969 B 2% No PCC 22,000
14 1959 A <1% No VIII 10,000
15 1946 B <1% No PCC 19,000
* Normal factor VIII and Ix levels are 50–150%
** Inhibitor indicates the presence of a circulating antibody against the deficient factor; for patients with hemophilia A, an inhibitor often necessitates the use of PCC instead of factor VIII concentrates to control bleeding.
Table 2 Transient HIV-1 Seroactivity in HIV-Exposed Hemophiliacs
Plasma Draw HIV-EIA HIV-Western blot IgG HIV-Western blot IgM **CCR5
Patient# Date 1S:C p17 p24 p66 gpl20 gpl60 p17 p24 p66 gpl20 gpl60 Deletion
1 82 0 - - - - - - - - - - +/-
83 1.065 - - - - - - - - - -
84 0 - - - - - - - - - -
86 0 - - - - - - - - - -
88 0 - - - - - - - - - -
2 81 0 - +* - - - - - + - - ND
83 0 - + - - - - - + - -
84 0 - + - - - - - + - -
86 1.143 - + - - - - - + - -
87 0 - + - - - - - + - -
89 0 - + - - - - - + - -
3 81 0 - + - - - - + - - - +/+
82 0 - +* - - - - + - - -
84 0 - + - - - - + - - -
86 0 - + - - - - + - - -
89 0 - ± - - - - + - - -
91 0 - ± - - - - + - - -
4 83 0 - + + + +* - + + + + +/+
85 0 - + - - - - - - - -
86 0 - + - - - - - - - -
89 0 - + - - - - - - - -
92 0 - + - - - - - - - -
5 80 0 - - - - - - - - - - +/+
81 2.437 - + + - - - - - - -
85 0 - + + - - - - - - -
86 0 - + + - - - - - - -
87 0 + + - + - - - - -
88 0 - - - - - - - - - -
90 0 - - - - - - - - - -
92 0 - - - - - - - - - -
6 81 0 - - - - - - - - - - +/+
82 0 - - - - - - - - - -
85 1.519 - - - - - - - - - -
86 1.104 - - - - - - - - - -
87 1.117 - - - - - - - - - -
88 0 - - - - - - - - - -
7 82 1.143 - - - - - - - - - - +/+
83 0 - - - - - - - - - -
86 0 - - - - - - - - - -
8 82 0 - + - - - +* + - - - +/+
84 0 - + - - - + + - - -
85 0 - + - - - + + - - -
86 0 - + - - - ± ± - - -
92 0 - + - - - ± ± - - -
9 80 0 - + - - - ± + - - - +/+
81 0 - + - - - + + - - -
82 0 - +* - - - +* + - - -
85 0 - + - - - + + - - -
87 0 - + - - - + + - - -
92 0 - + - - - + + - - -
10 81 0 - + - - - + + - - - +/+
84 0 - + - - - + + - - -
85 0 - + - - - + + - - -
86 0 - + - - - + + - - -
89 0 - + - - - + + - - -
90 0 - + - - - + + - - -
92 0 - + - - - + + - - -
11 85 0 - - - - - - + - - - -/-
86 0 - - - - - - + - - -
87 0 - - - - - - + - - -
89 0 - - - - - - + - - -
12 83 0 - - ± - - - - - - - +/+
84 0 - - ± - - - - - - -
85 0 - - ± - - - - - - -
87 0 - - ± - - - - - -
91 0 - - ± - - - - - - -
13 81 0 - + - - - - ± - - - +/+
82 0 - + - - - - ± - - -
83 0 - + - - - - ± - - -
84 0 - + - - - - ± - - -
85 0 - + - - - - ± - - -
87 0 - + - - - - ± - - -
91 0 - + - - - - ± - - -
14 80 0 - - - - - - - - - - +/-
82 0 - - - - - - - - - -
84 0 - - - - - - - - - -
85 0 - - - - - - - - - -
86 0 - - - - - - - - - -
92 0 - - - - - - - - - -
15 83 0 - - - - - - - - - - +/+
84 0 - - - - - - - - - -
85 0 - - - - - - - - - -
86 0 - - - - - - - - - -
91 0 - - - - - - - - - -
1S:C Denotes the ratio of sample absorbance value to cut-off value.
*Sample displaying the most reactivity in archived plasma set with fluctuating anti-HIV-l antibody reactivity on WB analysis.
**CCR5 Deletion Analysis
ND = not done
+/+ = wildtype,
+/- = mutant heterozygous
-/- = mutant homozygous
Assessment of current HIV-1-proviral DNA status
Post-1990 peripheral blood mononuclear cells were assessed for the presence of HIV-1 proviral DNA in all patients whose samples showed anti-HIV-1 reactivity using HIV-1-PCR analysis (kindly performed in a blinded manner using appropriate positive and negative controls by Charles Schable at the Centers for Disease Control and Prevention, Atlanta, GA.) All currently seronegative patients were found to be negative for HIV-1 proviral DNA by PCR analysis (data not shown).
Passive HIV-1 reactive antibodies in clotting Factor concentrates
It was possible that the HIV-1 activity that we detected results from the presence of passive anti-HIV antibodies that were present in a recently transfused CFC just prior to the drawing of the plasma sample we tested and therefore could be detectable in the plasma sample we tested. To assess the feasibility that the HIV-1 reactive antibodies present in contaminated CFCs could passively give rise to false-positive results in our testing, we reconstituted four factor-VIII concentrates and one PCC that were manufactured between 1981 and 1984 and tested them for the presence of anti-HIV-1 reactive antibodies. None of the reconstituted concentrates displayed reactivity on HIV-EIA, although a control aliquot of reconstituted factor spiked with HIV positive serum was reactive (data not shown) suggesting that the observed HIV-1 reactivity on our assays was not the result of the passive presence of HIV-1 reactive antibodies in CFCs. The reconstituted PCC was negative by HIV-1 WB analysis, however, the reconstituted factor VIII lots were found to have HIV-1 Wb reactivity at a dilution of 1:50 for the viral antigens p17, p24 and gpl60. One reconstituted factor HIV-1 sample had detectable anti-HIV-1 reactivity at a dilution of 1:100 (which was the dilution used for the hemophiliac plasma samples) but not beyond. No HIV-1 EIA or Wb reactivity was detected in clotting factor concentrates manufactured after 1985 (data not shown).
Presence of CCR-5 deletion mutations
To determine whether hetero- or homozygous CCR-5 deletion mutations were present in these patients, polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analyses were performed on samples obtained from all patients except patient 2 (Table 2). As expected most patients (11/14) were homozygous for wild type CCR -5. 2 of 14 patients were heterozygous for CCR-5Δ32. Interestingly, patient 11, who was homozygous positive for the mutation, was never infected; however, he was the most heavily exposed to the most infectious concentrate (factor VIII) and was an IV drug user. Because our laboratory has been following longitudinally a number of hemophiliacs (113) who have been exposed to CFCs, including the 15 patients followed throughout this study, we determined the CCR5 genotypes of these individuals to establish the frequency of CCR5 mutations. Expectedly, the vast majority of the hemophiliacs tested (113) were wild-type for CCR5; however, 14/131 (10.7%) and 2/131 (1.5%) were heterozygous and homozygous mutants of CCR5, respectively. Of these patients, 47 were HIV-1-positive, where 41/47 (87.2%), 6/47 (12.8%) and 0/47 (0%) were homozygous wild-type, heterozygous and homozygous mutant for CCR5, in that order. Of the remaining patients that were HIV-1-negative genotyped (a total of 84), 74/84 (88.1%), 8/84 (9.5%) and 2/84 (2.4%) were homozygous wild-type, heterozygous and homozygous mutant for CCR5 respectively.
Discussion
The presence of antibodies with specificity for multiple HIV-1 proteins is one of the diagnostic hallmarks of infection with the HIV-1 virus. In virtually all patients infected with HIV-1 permanent anti-viral antibodies are detectable typically within 3 to 12 weeks following exposure to the virus, and continue to increase in titer during the early phase of infection. Following this early phase, anti-HIV-1 antibody titer generally remains constant until the end stages of AIDS when fluctuations in total anti-HIV immunoglobulin may occur. Several studies have attempted to evaluate the existence of individuals who have been at high-risk of exposure to presumably infectious HIV, but who have resisted HIV-1 infection. This work has focused on several populations of high-risk HIV-1 seronegative individuals including intravenous drug users [24-26], HIV-1 exposed health care workers [27-29], sexual partners of HIV-1 infected individuals [27,30-40], female sex workers who have engaged in repeated unprotected sex with HIV-1 infected partners [30-32,38-44], HIV-1 uninfected children born from HIV-1 infected mothers [31,39,40,45,46] and hemophiliacs with a high probability of HIV-1 exposure from contaminated clotting factor concentrates [7,47-50].
In this study we retrospectively identified transient anti-HIV-1 antibody reactivity in archived plasma sets from currently HIV seronegative hemophiliacs who had a high probability of intravenous exposure to HIV contaminated CFCs. To accomplish this we used diagnostic methods for the detection of HIV reactive antibodies that are substantially more sensitive than earlier versions that were first introduction in the mid-80s when many of our plasma samples were originally tested and found to be negative. Confirmation of the current negative HIV status of the hemophiliacs whose archived samples were analyzed in this study was accomplished using HIV-l PCR analysis on recently obtained PBMC from these individuals.
Using HIV-EIA analysis we found that 5/15 (33%) archived plasma sets contained samples transiently reactive at one or more time point in 1987 or before. HIV-EIA reactivity above baseline is seen in less than 0.2% of healthy blood donors with no known exposure to HIV (HIV-1/HIV-2 EIA packet insert, Abbott Laboratories). One interpretation of these results is that there was a temporary appearance of low level antibodies reactive with HIV antigens in some hemophiliacs during, or shortly after, the most likely time of exposure to infectious or non-infectious HIV-1 in contaminated CFCs (1980–1985). The level of reactivity that we detected on the HIV-EIA was clearly above the baseline cut-off values determined by the respective controls for the assay, although lower than that typically seen in HIV-1 seropositive individuals.
None of the samples that were reactive on the HIV EIA showed reactivity against HIV-2 proteins by WB analysis suggesting that the HIV-2 antigens present on the EIA assay (Abbott Laboratories) were not likely responsible for the observed reactivity. The lack of anti-HIV-1 reactivity in samples older than those displaying reactivity suggested that these observations were not an artefact of the prolonged frozen storage of the plasma (Table 2, patient 1, 2 and 4–7).
Of the archived plasma sets that we tested, 10/15 (75%) had detectable IgG and/or IgM antibodies reactive with one or more HIV antigen on HIV-Wb analysis, which was most commonly directed against the p24 group antigen (figure 1 and table 2). Although most of the Wb reactivity that we observed would be classified as HIV-1 indeterminate, one archived plasma sample set had fluctuating weak IgG reactivity against multiple HIV proteins which included p24, p66 and gp160. This individual was also reactive by HIV-EIA analysis but only in one year corresponding to Wb activity (table 2, patient 5). A second hemophiliac, not reactive by HIV-EIA analysis, also had plasma antibodies reactive to HIV p24, p66, gp120 and gp160 antigens (Fig. 1, panel 4 and Table 2 patient 4). The p24 reactivity observed in this archived plasma set was consistently present in all of samples, however, the p66, gp120 and gp160 reactivity was observed only in the 1983 sample. Fluctuating anti-HIV reactive antibody titer was noted in samples from 5/10 (50%) of the archived plasma sets (Fig. 1 strip set 2–4 and Table 2, asterisks by patients 2–4, 8 and 9).
Indeterminate HIV-1 WB reactivity has been typically detected in only 5–7% of healthy EIA negative controls [51]. Although indeterminate results occurring in up to 32% of low-risk healthy populations have been reported with approximately half of these being attributable to p24 reactivity [52]. However, in our study, we observed that 10/15 (67%) had indeterminate reactivity in the archived plasma sets that we tested from hemophiliacs exposed to HIV-1 contaminated CFCs. Not surprisingly, reactivity to the p24 group antigen was the most frequent pattern that we observed. Most individuals when exposed to infectious HIV-1 will develop anti-gp120/160 envelope antibodies in addition to the p24 core antigen. In contrast, exposure to lysed HIV virus results almost exclusively in an anti-p24 response (Steve Alexander, personnel observations). Based on the pattern of HIV-Wb reactivity observed in the archived plasma sets tested in this study, our population of hemophiliacs were likely exposed predominantly to inactivated HIV-1. It is also possible that patterns of Wb reactivity more consistent with exposure to infectious HlV-1 may have existed in the hemophiliacs that we studied but was at a time not available in our archived collection of samples.
Among the five archived plasma sets that contained samples displaying HIV-EIA reactivity, only two (40%) also displayed HIV-Wb reactivity, which was not necessarily in corresponding years (Table 2., patients 2 and 4). Discrepancies between these two assays have previously been noted [53,54]. The HIV-Wb assay is an extremely sensitive method for the detection of antibodies which recognize predominantly, if not exclusively linearized epitopes. We have typically been able to detect anti-HIV serum antibodies from infected individuals when diluted more than a million fold. It is possible that the low level of reactivity noted in our archived plasma sets on HIV-Wb analysis may frequently have been beneath the limit of delectability by HIV-EIA. Discrepancies in reactivity between these two immunodiagnostic methods may also have resulted from the potential availability of conformational viral epitopes present on the HIV-EIA but not the HIV-Wb.
To determine whether archived plasma reactivity to HIV proteins could have been the result of passive acquired anti-HIV antibody contained in clotting factor concentrate, we assayed several concentrates made between 1981 and 1984 for the presence of anti-HIV-1 activity. One factor VIII concentrate from each of the 4 different U.S. manufacturers and one PCC made by the manufacturer that supplied over 90% of the PCCs used by our patient population were analyzed. Treatment with reconstituted clotting factor concentrate resulted in an approximate 1:100 dilution upon infusion into the blood stream (50 ml of reconstituted concentrate into 5 liters of whole blood). Accordingly, clotting factor concentrates were first diluted 1:100 prior to being run at the standard dilution for analysis (1:100 for HIV-Wb and 1:1.25 for the HIV-EIA). Under these parameters no anti-HIV reactivity was detectable in any of the clotting factor concentrates that were tested (data not shown). Additional studies in which clotting factor concentrates were spiked with limiting dilutions of HIV-1 specific antibodies were reactive indicating that the failure to detect anti-HIV antibodies was not due to interference by other clotting factor concentrate components (data not shown). Using more concentrated reconstituted clotting factor concentrates (a total dilution of 1:50) we could detect some HIV-Wb reactivity in all four of the factor VIII concentrates tested. This reactivity was with the p17, p24 and gp160 viral antigens, which confirms that CFCs were contaminated with blood products from HIV-seropositive donors (data not shown). Reactivity to the p17 and p160 viral proteins was rarely observed in the indeterminate samples from our archived plasma sets. This pattern of HIV-Wb reactivity would have been better represented in our archived plasma sets if the observed reactivity resulted from passively acquired anti-HIV antibodies present in contaminated clotting factor concentrates and introduced to the patient at the time of CFC transfusion. HIV-1-negative individuals (2.4%) had a slightly higher incidence of homozygosity for the deletion mutation of CCR5 compared with HIV-1-positive subjects (0%). These values reflect essentially those observed in the normal population as well as those in large cohort hemophiliac populations published earlier [11-14,16]. In this study there was no apparent protective advantage of CCR-5Δ32 heterozygosity in terms of HIV infection as has been reported previously [15-18].
Conclusion
Our results suggest that some severe hemophiliacs who were repeatedly exposed to CFCs contaminated with infectious and/or non-infectious HIV-1, immunologically processed some of the viral antigens but were not infected. It is expected that some of the HIV proteins in CFCs that the hemophiliacs in our study were exposed to would have been associated with non-infectious particles. However, we feel it likely that some of these individuals were transiently infected with HIV and then cleared the infection. The anti-HIV humoral reactivity that we detected would appear to be insufficient to abort a viral infection but the lack of any archived PBMC make it impossible to assess the degree to which any of our hemophiliacs may have mounted an anti-HIV cellular immune response at the time of exposure. It is also possible that some of the hemophiliacs in our study may have been exposed to an immunizing, but not infectious doses of HIV.
Methods
Patients
Archived plasma samples from 15 HIV seronegative individuals with either hemophilia A or B were selected for this study. All had been regularly followed by the Louisiana Comprehensive Hemophilia Care Center at Tulane University School of Medicine. Patient characteristics and clotting factor concentrate exposure history are given in table 1. Patients were chosen for study based upon use of clotting factor concentrates in excess of 5000 units per year from 1979 to 1985 and on availability of archived plasma samples for retrospective testing. At initial testing with first generation HIV enzyme linked immunosorbent assays (ELISA) in 1985 all of the individuals included in the present study were categorized as HIV seronegative. All studies described below were performed on samples of citrated plasma that had been stored at -70°C.
Detection of antibodies to HIV
HIV-1AB HIV-1IHIV-2 (rDNA) EIA
Archived plasma samples were reassessed for the presence of plasma antibodies reactive with HIV-1 and/or HIV-2 antigens using the HIV-1AB HIV-1IHIV-2 (rDNA) EIA kit (HIV-BIA, Abbott Laboratories) and HIV-1 and HIV-2 Wb assays (Cambridge Biotech, Rockville, Maryland) according to the manufactures protocol. In brief, the HIV-EIA is a current generation ELISA (1992) that utilizes a polystyrene bead coated with recombinant HIV-1 Env and Gag peptides and HIV-2 Env peptides. Test or control plasma was incubated at a dilution of 1:1.25 with viral antigen coated beads. Following washing, HIV reactive antibodies were detected by incubation of the bead-antibody complex with horseradish-peroxidase labeled HIV-1 and HIV-2 peptides that bind to all available open F'Ab sites. The enzyme-peptide-antibody complex was detected using a colorometric developing substrate solution comprised of 0-phenylene diamine~ 2HCl and analyzed spectrophotometrically at a wavelength of 492 nm. The frequency of reactivity on the HIV-EIA when tested on random blood donors is 0.16%. (HIV-1AB HIV-1/HIV-2 (rDNA) EIA, manual 83-8291IR4, 1992. Abbott Laboratories).
HIV-1 and HIV-2 Wb
HIV-1 and HIV-2 Wb analysis was performed using nitrocellulose strips containing electrophoretically separated and transferred proteins from inactivated HIV-1 or HIV-2 lysates. HIV-1 and HIV-2 Wb strips were subsequently incubated with test or control plasma at a dilution of 1:100. HIV reactive antibodies were visualized using biotinylated goat-anti human IgG and IgM, avidin-conjugated horseradish peroxidase and the colorimetric substrates 4-chloro-l-naphthol. Results on the HIV-1 Wb were classified as negative if no bands are present, and positive if any two or more of the following bands were present; p24, gp4l, gp120 and gp160 and had a reactivity score equal to or greater than the weak positive control. Indeterminate classification was given when banding was present, but did not meet the criteria for a positive interpretation [55]. The frequency of seropositivity on the HIV-l Wb when tested on random blood donors is 0.15% (HIV-1 Cambridge Biotech HIV-1 Western blot kit package insert).
Passive HIV reactive antibodies in clotting factor concentrates
We assayed four factor VIII and one prothrombin complex concentrate (PCC) each manufactured between 1981 and 1984 for the presence of HIV reactive antibodies. All concentrates had been stored at 4°C in their original sealed vials in a lyophilized state. Each was reconstituted for the present study according to the manufacturer's instructions using sterile water. Despite their age, all were readily reconstituted and had a normal appearance. Reconstituted concentrates were immediately aliquoted into l ml cryovials and placed at -70°C until their use [56]. To assess HIV reactivity, reconstituted factor concentrates were diluted 1:1.33, 1:10, 1:50, and 1:250 for HIV-1/HIV-2 RIA analysis and 1:50, 1:100, 1:200, 1:400, 1:1,000, 1:5,000, and 1:10,000 for HIV-1 Wb analysis. To determine the degree to which other components in clotting factor concentrates may interfere with the EIA or the Wb, a factor concentrate positive control was made by spiking aliquots of reconstituted clotting factor concentrates manufactured post 1990 with HIV-1 positive control sera and assayed for reduced reactivity.
Determination of CCR-5 Genotype by PCR-RFLP
Whole blood was lysed in RBC lysis solution (DNA isolation kit, Gentra Systems) for 1 minute at room temperature. Lysates were then centrifuged for 20 seconds at 13,000–16,000*g and the supernant was removed. Cell pellets were vortexed in residual liquid and 300 μl of cell lysis solution was added. CCR-5 genotyping was determined using PCR-RFLP as has been previously described.
Confirmation of HIV infection status
Due to the current HIV seronegative status of the hemophiliac population that we retrospectively analyzed in this study, it was considered unlikely that any were currently sub-clinically infected with HIV. However, to confirm this, recent peripheral blood mononuclear cell samples were assessed for the presence of HIV proviral DNA using HIV-1 PCR analysis.
Competing interests
The author(s) declare that they have no competing interests
Authors' contributions
SAT designed and performed the EIA and Wb experiments, analyzed the data and wrote the manuscript. Chemokine receptor PCR-RFLP was performed and analyzed by CAM and HEM. SSA provided intellectual assistance with Wb interpretation. RFG and CAL oversaw the design, development, implementation and analysis of the data for the project and edited the manuscript.
Acknowledgements
The authors wish to acknowledge the efforts of Ann Meyer M.T., A.S.C.P. in collecting and organizing archived plasma samples and to thank Ming Lu for technical assistance with PCR-RFLP assays. This work was supported in part by a Judith Graham Pool Postdoctoral Fellowship to SAT from The National Hemophilia Foundation. RFG was supported by NIH grants M25904, M34754 and DE10862. CAL was supported by NIH grant HM6670.
==== Refs
Rein B Wuiff K Andes A Hemophilia and HIV infection in Louisiana Louisiana Morbidity Report 1988
National Hemophilia Foundation HIV seroprevalence in the Hemophilia Community. Hemophilia Information Exchange Medical Bulletin 1991 July 19
Kroner BL Goedert JJ Blattner WA Wilson SE Carrington MN Mann DL Concordance of human leukocyte antigen haplotype-sharing, CD4 decline and AIDS in hemophilic siblings. Multicenter Hemophilia Cohort and Hemophilia Growth and Development Studies Aids 1995 9 275 280 7755916
Kroner BL Rosenberg PS Aledort LM Alvord WG Goedert JJ HIV-1 infection incidence among persons with hemophilia in the United States and western Europe, 1978–1990. Multicenter Hemophilia Cohort Study J Acquir Immune Defic Syndr 1994 7 279 286 8106967
Goedert JJ Cohen AR Kessler CM Eichinger S Seremetis SV Rabkin CS Yellin FJ Rosenberg PS Aledort LM Risks of immunodeficiency, AIDS, and death related to purity of factor VIII concentrate. Multicenter Hemophilia Cohort Study Lancet 1994 344 791 792 7916077 10.1016/S0140-6736(94)92345-0
Goedert JJ Garvey L Hilgartner MW Blatt PM Aledort LM Cohen AR Kessler CM White GC Mandalaki T Cook RA Risk of HIV infection and AIDS in women and girls with coagulation disorders Aids 1994 8 564 565 8011269
Ludlam CA Tucker J Steel CM Tedder RS Cheingsong-Popov R Weiss RA McClelland DB Philp I Prescott RJ Human T-lymphotropic virus type III (HTLV-III) infection in seronegative haemophiliacs after transfusion of factor VIII Lancet 1985 2 233 236 2862416 10.1016/S0140-6736(85)90288-0
Fabio G Marchini M Scorza Smeraldi R Zarantonello M Tarantini G Gringeri A Antonioli R Possible association of HLA-DR2 phenotype and detectable human immunodeficiency virus (HIV) p24 antigen in HIV-positive patients J Infect Dis 1993 167 499 500 8421192
Papasteriades C Varla M Economidou J Marcacis K Mitsouli C Louisou K Mandalaki T Roumeliotou A Papaevangelou G High frequency of HLA-DR5 in Greek patients with haemophilia A and haemophilia B Tissue Antigens 1986 28 84 87 3018962
Steel CM Ludlam CA Beatson D Peutherer JF Cuthbert RJ Simmonds P Morrison H Jones M HLA haplotype A1 B8 DR3 as a risk factor for HIV-related disease Lancet 1988 1 1185 1188 2897006 10.1016/S0140-6736(88)92009-0
Dean M Carrington M Winkler C Huttley GA Smith MW Allikmets R Goedert JJ Buchbinder SP Vittinghoff E Gomperts E Genetic restriction of HIV-1 infection and progression to AIDS by a deletion allele of the CKR5 structural gene. Hemophilia Growth and Development Study, Multicenter AIDS Cohort Study, Multicenter Hemophilia Cohort Study, San Francisco City Cohort, ALIVE Study Science 1996 273 1856 1862 8791590
Liu R Paxton WA Choe S Ceradini D Martin SR Horuk R MacDonald ME Stuhlmann H Koup RA Landau NR Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection Cell 1996 86 367 377 8756719 10.1016/S0092-8674(00)80110-5
Samson M Libert F Doranz BJ Rucker J Liesnard C Farber CM Saragosti S Lapoumeroulie C Cognaux J Forceille C Resistance to HIV-1 infection in caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene Nature 1996 382 722 725 8751444 10.1038/382722a0
Smith MW Dean M Carrington M Winkler C Huttley GA Lomb DA Goedert JJ O'Brien TR Jacobson LP Kaslow R Contrasting genetic influence of CCR2 and CCR5 variants on HIV-1 infection and disease progression. Hemophilia Growth and Development Study (HGDS), Multicenter AIDS Cohort Study (MACS), Multicenter Hemophilia Cohort Study (MHCS), San Francisco City Cohort (SFCC), ALIVE Study Science 1997 277 959 965 9252328 10.1126/science.277.5328.959
Kantor R Barzilai A Varon D Martinowitz U Gershoni JM Prevalence of a CCR5 gene 32-bp deletion in an Israeli cohort of HIV-1-infected and uninfected hemophilia patients J Hum Virol 1998 1 299 301 10195255
Kupfer B Kaiser R Brackmann HH Effenberger W Rockstroh JK Matz B Schneweis KE Protection against parenteral HIV-1 infection by homozygous deletion in the C-C chemokine receptor 5 gene Aids 1999 13 1025 1028 10397531 10.1097/00002030-199906180-00004
Pasi KJ Sabin CA Jenkins PV Devereux HL Ononye C Lee CA The effects of the 32-bp CCR-5 deletion on HIV transmission and HIV disease progression in individuals with haemophilia Br J Haematol 2000 111 136 142 11091193 10.1046/j.1365-2141.2000.02325.x
Zamarchi R Indraccolo S Minuzzo S Coppola V Gringeri A Santagostino E Vicenzi E De Silvestro G Biagiotti R Baldassarre C Frequency of a mutated CCR-5 allele (delta32) among Italian healthy donors and individuals at risk of parenteral HIV infection AIDS Res Hum Retroviruses 1999 15 337 344 10082117 10.1089/088922299311303
Agius G Biggar RJ Alexander SS Waters DJ Drummond JE Murphy EL Weiss SH Levine PH Blattner WA Human T lymphotropic virus type I antibody patterns: evidence of difference by age and risk group J Infect Dis 1988 158 1235 1244 3198937
Burke DS Redfield RR Putman P Alexander SS Variations in Western blot banding patterns of human T-cell lymphotropic virus type III/lymphadenopathy-associated virus J Clin Microbiol 1987 25 81 84 3540002
Dock NL Lamberson HV JrO'Brien TA Tribe DE Alexander SS Poiesz BJ Evaluation of atypical human immunodeficiency virus immunoblot reactivity in blood donors Transfusion 1988 28 412 418 3420668 10.1046/j.1537-2995.1988.28588337326.x
Farzadegan H Polis MA Wolinsky SM Rinaldo CR JrSninsky JJ Kwok S Griffith RL Kaslow RA Phair JP Polk BF Loss of human immunodeficiency virus type 1 (HIV-1) antibodies with evidence of viral infection in asymptomatic homosexual men. A report from the Multicenter AIDS Cohort Study Ann Intern Med 1988 108 785 790 3285743
Farzadegan H Taylor E Hardy W Odaka N Polk BF Performance of serological assays for early detection of human immunodeficiency virus type 1 seroconversion J Clin Microbiol 1989 27 1882 1884 2671037
Beretta A Weiss SH Rappocciolo G Mayur R De Santis C Quirinale J Cosma A Robbioni P Shearer GM Berzofsky JA Human immunodeficiency virus type 1 (HIV-1)-seronegative injection drug users at risk for HIV exposure have antibodies to HLA class I antigens and T cells specific for HIV envelope J Infect Dis 1996 173 472 476 8568315
Boiocchi M Franceschi S Serraino D Guarneri S Tavian M Diodato S Persistent lack of human immunodeficiency virus infection in intravenous drug users at very high risk J Acquir Immune Defic Syndr 1993 6 633 635 8496794
Moore JD Cone EJ Alexander SS Jr HTLV-III seropositivity in 1971–1972 parenteral drug abusers--a case of false positives for evidence of viral exposure? N Engl J Med 1986 314 1387 1388 3010112
Clerici M Giorgi JV Chou CC Gudeman VK Zack JA Gupta P Ho HN Nishanian PG Berzofsky JA Shearer GM Cell-mediated immune response to human immunodeficiency virus (HIV) type 1 in seronegative homosexual men with recent sexual exposure to HIV-1 J Infect Dis 1992 165 1012 1019 1533867
Pinto LA Landay AL Berzofsky JA Kessler HA Shearer GM Immune response to human immunodeficiency virus (HIV) in healthcare workers occupationally exposed to HIVcontaminated blood Am J Med 1997 102 21 24 9845492 10.1016/S0002-9343(97)00056-9
Pinto LA Sullivan J Berzofsky JA Clerici M Kessler HA Landay AL Shearer GM ENVspecific cytotoxic T lymphocyte responses in HIV seronegative health care workers occupationally exposed to HIV-contaminated body fluids J Clin Invest 1995 96 867 876 7635981
Detels R Liu Z Hennessey K Kan J Visscher BR Taylor JM Hoover DR Rinaldo CR JrPhair JP Saah AJ Resistance to HIV-1 infection. Multicenter AIDS Cohort Study J Acquir Immune Defic Syndr 1994 7 1263 1269 7965637
Detels R Mann D Carrington M Hennessey K Wu Z Hirji KF Wiley D Visscher BR Giorgi JV Resistance to HIV infection may be genetically mediated Aids 1996 10 102 104 8924238
Haynes BF Pantaleo G Fauci AS Toward an understanding of the correlates of protective immunity to HIV infection Science 1996 271 324 328 8553066
Kelker HC Seidlin M Vogler M Valentine FT Lymphocytes from some long-term seronegative heterosexual partners of HIV-infected individuals proliferate in response to HIV antigens AIDS Res Hum Retroviruses 1992 8 1355 1359 1466956
Langlade-Demoyen P Ngo-Giang-Huong N Ferchal F Oksenhendler E Human immunodeficiency virus (HIV) nef-specific cytotoxic T lymphocytes in noninfected heterosexual contact of HIV-infected patients J Clin Invest 1994 93 1293 1297 8132769
O'Brien TR Blattner WA Waters D Eyster E Hilgartner MW Cohen AR Luban N Hatzakis A Aledort LM Rosenberg PS Serum HIV-1 RNA levels and time to development of AIDS in the Multicenter Hemophilia Cohort Study Jama 1996 276 105 110 8656501 10.1001/jama.276.2.105
O'Brien TR Padian NS Hodge T Goedert JJ O'Brien SJ Carrington M CCR-5 genotype and sexual transmission of HIV-1 Aids 1998 12 444 445 9520179
O'Brien TR VanDevanter N Paxton H Polan C Holmberg SD CD4+ T-lymphocyte counts among seronegative heterosexual partners of persons infected with human immunodeficiency virus type 1 J Acquir Immune Defic Syndr 1993 6 1374 1375 7902867
Rowland-Jones SL Dong T Fowke KR Kimani J Krausa P Newell H Blanchard T Ariyoshi K Oyugi J Ngugi E Cytotoxic T cell responses to multiple conserved HIV epitopes in HIV-resistant prostitutes in Nairobi J Clin Invest 1998 102 1758 1765 9802890
Rowland-Jones SL McMichael A Immune responses in HIV-exposed seronegatives: have they repelled the virus? Curr Opin Immunol 1995 7 448 455 7495507 10.1016/0952-7915(95)80087-5
Rowland-Jones SL Nixon DF Aldhous MC Gotch F Ariyoshi K Hallam N Kroll JS Froebel K McMichael A HIV-specific cytotoxic T-cell activity in an HIV-exposed but uninfected infant Lancet 1993 341 860 861 8096564 10.1016/0140-6736(93)93063-7
Fowke KR Dong T Rowland-Jones SL Oyugi J Rutherford WJ Kimani J Krausa P Bwayo J Simonsen JN Shearer GM HIV type 1 resistance in Kenyan sex workers is not associated with altered cellular susceptibility to HIV type 1 infection or enhanced betachemokine production AIDS Res Hum Retroviruses 1998 14 1521 1530 9840285
Fowke KR Kaul R Rosenthal KL Oyugi J Kimani J Rutherford WJ Nagelkerke NJ Ball TB Bwayo JJ Simonsen JN HIV-1-specific cellular immune responses among HIV-1-resistant sex workers Immunol Cell Biol 2000 78 586 595 11114968 10.1046/j.1440-1711.2000.00944.x
Fowke KR Nagelkerke NJ Kimani J Simonsen JN Anzala AO Bwayo JJ MacDonald KS Ngugi EN Plummer FA Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya Lancet 1996 348 1347 1351 8918278 10.1016/S0140-6736(95)12269-2
Rubsamen-Waigmann H Maniar J Gerte S Brede HD Dietrich U Mahambre G Pfutzner A High proportion of HIV-2 and HIV-1/2 double-reactive sera in two Indian states, Maharashtra and Goa: first appearance of an HIV-2 epidemic along with an HIV-1 epidemic outside of Africa Zentralbl Bakteriol 1994 280 398 402 8167434
Bryson YJ Pang S Wei LS Dickover R Diagne A Chen IS Clearance of HIV infection in a perinatally infected infant N Engl J Med 1995 332 833 838 7870139 10.1056/NEJM199503303321301
Bryson YJ Pang S Wi L A child found to be HIV positive shortly after birth appears now to be clear of the infection Nurs Times 1995 91 11 7731834
Chelucci C Hassan HJ Gringeri A Macioce G Mariani G Santagostino E Testa U Vulcano F Mannucci PM Peschle C PCR analysis of HIV-1 sequences and differential immunological features in seronegative and seropositive haemophiliacs Br J Haematol 1992 81 558 567 1390244
Lederman MM Jackson JB Kroner BL White GC 3rdEyster ME Aledort LM Hilgartner MW Kessler CM Cohen AR Kiger KP Human immunodeficiency virus (HIV) type 1 infection status and in vitro susceptibility to HIV infection among high-risk HIV-1-seronegative hemophiliacs J Infect Dis 1995 172 228 231 7797917
Shoebridge GI Barone L Wing-Simpson A Bennetts BH Nightingale BN Hensley WJ Gatenby PA Assessment of HIV status using the polymerase chain reaction in antibodypositive patients and high-risk antibody-negative haemophiliacs Aids 1991 5 221 224 2031695
Shoebridge GI Gatenby PA Nightingale BN Barone L Wing-Simpson A Bennetts BH Hensley WJ Polymerase chain reaction testing of HIV-1 seronegative at-risk individuals Lancet 1990 336 180 181 1973496 10.1016/0140-6736(90)91702-C
Talal N Dauphinee MJ Dang H Alexander SS Hart DJ Garry RF Detection of serum antibodies to retroviral proteins in patients with primary Sjogren's syndrome (autoimmune exocrinopathy) Arthritis Rheum 1990 33 774 781 2363733
Midthun K Garrison L Clements ML Farzadegan H Fernie B Quinn T Frequency of indeterminate western blot tests in healthy adults at low risk for human immunodeficiency virus infection. The NIAID AIDS Vaccine Clinical Trails Network J Infect Dis 1990 162 1379 1382 2230270
Downie JC Howard R Bowcock B Cunningham AL HIV-1 antibody testing strategy: evaluation of ELISA screening and western blot profiles in a mixed low risk/high risk patient population J Virol Methods 1989 26 291 303 2695539 10.1016/0166-0934(89)90111-0
Portincasa P Conti G Zannino T Visalli S Chezzi C Radioimmune western blotting in comparison with conventional western blotting, second and third generation ELISA assays for the serodiagnosis of HIV-1 infection New Microbiol 1994 17 169 176 7968651
Interpretation and use of the Western blot assay for serodiagnosis of human immunodeficiency virus type 1 infections MMWR 1989 38 1 7
Leissinger CA Stuckey WJ Unbound immunoglobulins are a major source of error in the quantitation of platelet surface-bound immunoglobulin levels Transfusion 1992 32 157 161 1542921 10.1046/j.1537-2995.1992.32292180146.x
|
16107217
|
PMC1232868
|
CC BY
|
2021-01-04 16:39:01
|
no
|
Virol J. 2005 Aug 17; 2:65
|
utf-8
|
Virol J
| 2,005 |
10.1186/1743-422X-2-65
|
oa_comm
|
==== Front
Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-691611531810.1186/1743-422X-2-69ResearchChloroquine is a potent inhibitor of SARS coronavirus infection and spread Vincent Martin J [email protected] Eric [email protected] Suzanne [email protected] Bobbie R [email protected] Pierre E [email protected] Thomas G [email protected] Nabil G [email protected] Stuart T [email protected] Special Pathogens Brach, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, 1600 Clifton Road, Atlanta, Georgia, 30333, USA2 Laboratory of Biochemical Neuroendocrinology, Clinical Research Institute of Montreal, 110 Pine Ave West, Montreal, QCH2W1R7, Canada2005 22 8 2005 2 69 69 12 7 2005 22 8 2005 Copyright © 2005 Vincent et al; licensee BioMed Central Ltd.2005Vincent 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
Severe acute respiratory syndrome (SARS) is caused by a newly discovered coronavirus (SARS-CoV). No effective prophylactic or post-exposure therapy is currently available.
Results
We report, however, that chloroquine has strong antiviral effects on SARS-CoV infection of primate cells. These inhibitory effects are observed when the cells are treated with the drug either before or after exposure to the virus, suggesting both prophylactic and therapeutic advantage. In addition to the well-known functions of chloroquine such as elevations of endosomal pH, the drug appears to interfere with terminal glycosylation of the cellular receptor, angiotensin-converting enzyme 2. This may negatively influence the virus-receptor binding and abrogate the infection, with further ramifications by the elevation of vesicular pH, resulting in the inhibition of infection and spread of SARS CoV at clinically admissible concentrations.
Conclusion
Chloroquine is effective in preventing the spread of SARS CoV in cell culture. Favorable inhibition of virus spread was observed when the cells were either treated with chloroquine prior to or after SARS CoV infection. In addition, the indirect immunofluorescence assay described herein represents a simple and rapid method for screening SARS-CoV antiviral compounds.
severe acute respiratory syndrome coronaviruschloroquineinhibitiontherapy
==== Body
Background
Severe acute respiratory syndrome (SARS) is an emerging disease that was first reported in Guangdong Province, China, in late 2002. The disease rapidly spread to at least 30 countries within months of its first appearance, and concerted worldwide efforts led to the identification of the etiological agent as SARS coronavirus (SARS-CoV), a novel member of the family Coronaviridae [1]. Complete genome sequencing of SARS-CoV [2,3] confirmed that this pathogen is not closely related to any of the previously established coronavirus groups. Budding of the SARS-CoV occurs in the Golgi apparatus [4] and results in the incorporation of the envelope spike glycoprotein into the virion. The spike glycoprotein is a type I membrane protein that facilitates viral attachment to the cellular receptor and initiation of infection, and angiotensin-converting enzyme-2 (ACE2) has been identified as a functional cellular receptor of SARS-CoV [5]. We have recently shown that the processing of the spike protein was effected by furin-like convertases and that inhibition of this cleavage by a specific inhibitor abrogated cytopathicity and significantly reduced the virus titer of SARS-CoV [6].
Due to the severity of SARS-CoV infection, the potential for rapid spread of the disease, and the absence of proven effective and safe in vivo inhibitors of the virus, it is important to identify drugs that can effectively be used to treat or prevent potential SARS-CoV infections. Many novel therapeutic approaches have been evaluated in laboratory studies of SARS-CoV: notable among these approaches are those using siRNA [7], passive antibody transfer [8], DNA vaccination [9], vaccinia or parainfluenza virus expressing the spike protein [10,11], interferons [12,13], and monoclonal antibody to the S1-subunit of the spike glycoprotein that blocks receptor binding [14]. In this report, we describe the identification of chloroquine as an effective pre- and post-infection antiviral agent for SARS-CoV. Chloroquine, a 9-aminoquinoline that was identified in 1934, is a weak base that increases the pH of acidic vesicles. When added extracellularly, the non-protonated portion of chloroquine enters the cell, where it becomes protonated and concentrated in acidic, low-pH organelles, such as endosomes, Golgi vesicles, and lysosomes. Chloroquine can affect virus infection in many ways, and the antiviral effect depends in part on the extent to which the virus utilizes endosomes for entry. Chloroquine has been widely used to treat human diseases, such as malaria, amoebiosis, HIV, and autoimmune diseases, without significant detrimental side effects [15]. Together with data presented here, showing virus inhibition in cell culture by chloroquine doses compatible with patient treatment, these features suggest that further evaluation of chloroquine in animal models of SARS-CoV infection would be warranted as we progress toward finding effective antivirals for prevention or treatment of the disease.
Results
Preinfection chloroquine treatment renders Vero E6 cells refractory to SARS-CoV infection
In order to investigate if chloroquine might prevent SARS-CoV infection, permissive Vero E6 cells [1] were pretreated with various concentrations of chloroquine (0.1–10 μM) for 20–24 h prior to virus infection. Cells were then infected with SARS-CoV, and virus antigens were visualized by indirect immunofluorescence as described in Materials and Methods. Microscopic examination (Fig. 1A) of the control cells (untreated, infected) revealed extensive SARS-CoV-specific immunostaining of the monolayer. A dose-dependant decrease in virus antigen-positive cells was observed starting at 0.1 μM chloroquine, and concentrations of 10 μM completely abolished SARS-CoV infection. For quantitative purposes, we counted the number of cells stained positive from three random locations on a slide. The average number of positively stained control cells was scored as 100% and was compared with the number of positive cells observed under various chloroquine concentrations (Fig. 1B). Pretreatment with 0.1, 1, and 10 μM chloroquine reduced infectivity by 28%, 53%, and 100%, respectively. Reproducible results were obtained from three independent experiments. These data demonstrated that pretreatment of Vero E6 cells with chloroquine rendered these cells refractory to SARS-CoV infection.
Figure 1 Prophylactic effect of chloroquine. Vero E6 cells pre-treated with chloroquine for 20 hrs. Chloroquine-containing media were removed and the cells were washed with phosphate buffered saline before they were infected with SARS-CoV (0.5 multiplicity of infection) for 1 h. in the absence of chloroquine. Virus was then removed and the cells were maintained in Opti-MEM (Invitrogen) for 16–18 h in the absence of chloroquine. SARS-CoV antigens were stained with virus-specific HMAF, followed by FITC-conjugated secondary antibodies. (A) The concentration of chloroquine used is indicated on the top of each panel. (B) SARS-CoV antigen-positive cells at three random locations were captured by using a digital camera, the number of antigen-positive cells was determined, and the average inhibition was calculated. Percent inhibition was obtained by considering the untreated control as 0% inhibition. The vertical bars represent the range of SEM.
Postinfection chloroquine treatment is effective in preventing the spread of SARS-CoV infection
In order to investigate the antiviral properties of chloroquine on SARS-CoV after the initiation of infection, Vero E6 cells were infected with the virus and fresh medium supplemented with various concentrations of chloroquine was added immediately after virus adsorption. Infected cells were incubated for an additional 16–18 h, after which the presence of virus antigens was analyzed by indirect immunofluorescence analysis. When chloroquine was added after the initiation of infection, there was a dramatic dose-dependant decrease in the number of virus antigen-positive cells (Fig. 2A). As little as 0.1–1 μM chloroquine reduced the infection by 50% and up to 90–94% inhibition was observed with 33–100 μM concentrations (Fig. 2B). At concentrations of chloroquine in excess of 1 μM, only a small number of individual cells were initially infected, and the spread of the infection to adjacent cells was all but eliminated. A half-maximal inhibitory effect was estimated to occur at 4.4 ± 1.0 μM chloroquine (Fig. 2C). These data clearly show that addition of chloroquine can effectively reduce the establishment of infection and spread of SARS-CoV if the drug is added immediately following virus adsorption.
Figure 2 Post-infection chloroquine treatment reduces SARS-CoV infection and spread. Vero E6 cells were seeded and infected as described for Fig. 1 except that chloroquine was added only after virus adsorption. Cells were maintained in Opti-MEM (Invitrogen) containing chloroquine for 16–18 h, after which they were processed for immunofluorescence. (A) The concentration of chloroquine is indicated on the top. (B) Percent inhibition and SEM were calculated as in Fig. 1B. (C) The effective dose (ED50) was calculated using commercially available software (Grafit, version 4, Erithacus Software).
Electron microscopic analysis indicated the appearance of significant amounts of extracellular virus particles 5–6 h after infection [16]. Since we observed antiviral effects by chloroquine immediately after virus adsorption, we further extended the analysis by adding chloroquine 3 and 5 h after virus adsorption and examined for the presence of virus antigens after 20 h. We found that chloroquine was still significantly effective even when added 5 h after infection (Fig. 3); however, to obtain equivalent antiviral effect, a higher concentration of chloroquine was required if the drug was added 3 or 5 h after adsorption.
Figure 3 Timed post-infection treatment with chloroquine. This experiment is similar to that depicted in Fig. 2 except that cells were infected at 1 multiplicity of infection, and chloroquine (10, 25, and 50 μM) was added 3 or 5 h after infection.
Ammonium chloride inhibits SARS-CoV infection of Vero E6 cells
Since chloroquine inhibited SARS-CoV infection when added before or after infection, we hypothesized that another common lysosomotropic agent, NH4Cl, might also function in a similar manner. Ammonium chloride has been widely used in studies addressing endosome-mediated virus entry. Coincidently, NH4Cl was recently shown to reduce the transduction of pseudotype viruses decorated with SARS-CoV spike protein [17,18]. In an attempt to examine if NH4Cl functions similarly to chloroquine, we performed infection analyses in Vero E6 cells before (Fig. 4A) and after (Fig. 4B) they were treated with various concentrations of NH4Cl. In both cases, we observed a 93–99% inhibition with NH4Cl at ≥ 5 mM. These data indicated that NH4Cl (≥ 5 mM) and chloroquine (≥ 10 μM) are very effective in reducing SARS-CoV infection. These results suggest that effects of chloroquine and NH4Cl in controlling SARS CoV infection and spread might be mediated by similar mechanism(s).
Figure 4 NH4Cl inhibits SARS-CoV during pre or post infection treatment. NH4Cl was added to the cells either before (A) or after (B) infection, similar to what was done for chloroquine in Figs 1 and 2. Antigen-positive cells were counted, and the results were presented as in Fig. 1B.
Effect of chloroquine and NH4Cl on cell surface expression of ACE2
We performed additional experiments to elucidate the mechanism of SARS-CoV inhibition by chloroquine and NH4Cl. Since intra-vesicular acidic pH regulates cellular functions, including N-glycosylation trimming, cellular trafficking, and various enzymatic activities, it was of interest to characterize the effect of both drugs on the processing, glycosylation, and cellular sorting of SARS-CoV spike glycoprotein and its receptor, ACE2. Flow cytometry analysis was performed on Vero E6 cells that were either untreated or treated with highly effective anti-SARS-CoV concentrations of chloroquine or NH4Cl. The results revealed that neither drug caused a significant change in the levels of cell-surface ACE2, indicating that the observed inhibitory effects on SARS-CoV infection are not due to the lack of available cell-surface ACE2 (Fig. 5A). We next analyzed the molecular forms of endogenous ACE2 in untreated Vero E6 cells and in cells that were pre-incubated for 1 h with various concentrations of either NH4Cl (2.5–10 mM) or chloroquine (1 and 10 μM) and labeled with 35S-(Met) for 3 h in the presence or absence of the drugs (Fig. 5B and 5C). Under normal conditions, we observed two immunoreactive ACE2 forms, migrating at ~105 and ~113 kDa, respectively (Fig. 5B, lane 1). The ~105-kDa protein is endoglycosidase H sensitive, suggesting that it represents the endoplasmic reticulum (ER) localized form, whereas the ~113-kDa protein is endoglycosidase H resistant and represents the Golgi-modified form of ACE2 [19]. The specificity of the antibody was confirmed by displacing the immunoreactive protein bands with excess cold-soluble human recombinant ACE2 (+ rhACE2; Fig. 5B, lane 2). When we analyzed ACE2 forms in the presence of NH4Cl, a clear stepwise increase in the migration of the ~113-kDa protein was observed with increasing concentrations of NH4Cl, with a maximal effect observed at 10 mM NH4Cl, resulting in only the ER form of ACE2 being visible on the gel (Fig. 5B, compare lanes 3–5). This suggested that the trimming and/or terminal modifications of the N-glycosylated chains of ACE2 were affected by NH4Cl treatment. In addition, at 10 mM NH4Cl, the ER form of ACE2 migrated with slightly faster mobility, indicating that NH4Cl at that concentration might also affect core glycosylation. We also examined the terminal glycosylation status of ACE2 when the cells were treated with chloroquine (Fig. 5C). Similar to NH4Cl, a stepwise increase in the electrophoretic mobility of ACE2 was observed with increasing concentrations of chloroquine. At 25 μM chloroquine, the faster electrophoretic mobility of the Golgi-modified form of ACE2 was clearly evident. On the basis of the flow cytometry and immunoprecipitation analyses, it can be inferred that NH4Cl and chloroquine both impaired the terminal glycosylation of ACE2, while NH4Cl resulted in a more dramatic effect. Although ACE2 is expressed in similar quantities at the cell surface, the variations in its glycosylation status might render the ACE2-SARS-CoV interaction less efficient and inhibit virus entry when the cells are treated with NH4Cl and chloroquine.
Figure 5 Effect of lysomotropic agents on the cell-surface expression and biosynthesis of ACE2. (A) Vero E6 cells were cultured for 20 h in the absence (control) or presence of chloroquine (10 μM) or NH4Cl (20 mM). Cells were labeled with anti-ACE2 (grey histogram) or with a secondary antibody alone (white histogram). (B) Biosynthesis of ACE2 in untreated cells or in cells treated with NH4Cl. Vero E6 cells were pulse-labeled for 3 h with 35S-Met, and the cell lysates were immunoprecipitated with an ACE2 antibody (lane 1). Preincunbation of the antibody with recombinant human ACE2 (rhACE2) completely abolished the signal (lane 2). The positions of the endoglycosidase H-sensitive ER form and the endoglycosidase H-resistant Golgi form of ACE2 are emphasized. Note that the increasing concentration of NH4Cl resulting in the decrease of the Golgi form of ACE2. (C) A similar experiment was performed in the presence of the indicated concentrations of chloroquine. Note the loss of terminal glycans with increasing concentrations of chloroquine. (D) The terminal glycosidic modification of ACE2 was evaluated by neuraminidase treatment of immunoprecipitated ACE2. Here cells were treated with 1–25 μM concentrations of chloroquine during starvation, pulse, and 3-h chase.
To confirm that ACE2 undergoes terminal sugar modifications and that the terminal glycosylation is affected by NH4Cl or chloroquine treatment, we performed immunopreipitation of 35S-labeled ACE2 and subjected the immunoprecipitates to neuraminidase digestion. Proteins were resolved using SDS-PAGE (Fig 5D). It is evident from the slightly faster mobility of the Golgi form of ACE2 after neuraminidase treatment (Fig 5D, compare lanes 1 and 2), that ACE2 undergoes terminal glycosylation; however, the ER form of ACE2 was not affected by neuraminidase. Cells treated with 10 μM chloroquine did not result in a significant shift; whereas 25 μM chloroquine caused the Golgi form of ACE2 to resolve similar to the neuraminidase-treated ACE2 (Fig 5D, compare lanes 5 and 6). These data provide evidence that ACE2 undergoes terminal glycosylation and that chloroquine at anti-SARS-CoV concentrations abrogates the process.
Effect of chloroquine and NH4Cl on the biosynthesis and processing of SARS-CoV spike protein
We next addressed whether the lysosomotropic drugs (NH4Cl and chloroquine) affect the biosynthesis, glycosylation, and/or trafficking of the SARS-CoV spike glycoprotein. For this purpose, Vero E6 cells were infected with SARS-CoV for 18 h. Chloroquine or ammonium chloride was added to these cells during while they were being starved (1 h), labeled (30 min) or chased (3 h). The cell lysates were analyzed by immunoprecipitation with the SARS-specific polyclonal antibody (HMAF). The 30-min pulse results indicated that pro-spike (proS) was synthesized as a ~190-kDa precursor (proS-ER) and processed into ~125-, ~105-, and ~80-kDa proteins (Fig. 6A, lane 2), a result identical to that in our previous analysis [6]. Except for the 100 μM chloroquine (Fig. 6A, lane 3), there was no significant difference in the biosynthesis or processing of the virus spike protein in untreated or chloroquine-treated cells (Fig. 6A, lanes 4–6). It should be noted that chloroquine at 100 μM resulted in an overall decrease in biosynthesis and in the levels of processed virus glycoprotein. In view of the lack of reduction in the biosynthesis and processing of the spike glycoprotein in the presence of chloroquine concentrations (10 and 50 μM) that caused large reductions in SARS-CoV replication and spread, we conclude that the antiviral effect is probably not due to alteration of virus glycoprotein biosynthesis and processing. Similar analyses were performed with NH4Cl, and the data suggested that the biosynthesis and processing of the spike protein were also not negatively affected by NH4Cl (Fig. 6A, lanes 7–12). Consistent with our previous analysis [6], we observed the presence of a larger protein, which is referred to here as oligomers. Recently, Song et al. [20] provided evidence that these are homotrimers of the SARS-CoV spike protein and were incorporated into the virions. Interestingly, the levels of the homotrimers in cells treated with 100 μM chloroquine and 40 and 20 mM NH4Cl (Fig. 6A, lanes 3, 9, and 10) were slightly lower than in control cells or cells treated with lower drug concentrations.
Figure 6 Effects of NH4Cl and chloroquine (CQ) on the biosynthesis, processing, and glycosylation of SARS-CoV spike protein. Vero E6 cells were infected with SARS-CoV as described in Fig. 2. CQ or NH4Cl was added during the periods of starvation (1 h) and labeling (30 min) with 35S-Cys and followed by chase for 3 h in the presence of unlabeled medium. Cells were lysed in RIPA buffer and immunoprecipitated with HMAF. Virus proteins were resolved using 3–8% NuPAGE gel (Invitrogen). The cells presented were labeled for 30 min (A) and chased for 3 h (B). The migration positions of the various spike molecular forms are indicated at the right side, and those of the molecular standards are shown to the left side. proS-ER and proS-Golgi are the pro-spike of SARS-Co in the ER and Golgi compartments, respectively and proS-ungly is the unglycosylated pro-spike ER.
The data obtained from a 30-min pulse followed by a 3-h chase (Fig. 6B, lanes 2 and 8) confirmed our earlier observation that the SARS-CoV spike protein precursor (proS-ER) acquires Golgi-specific modifications (proS-Golgi) resulting in a ~210-kDa protein [6]. Chloroquine at 10, 25, and 50 μM had no substantial negative impact on the appearance of the Golgi form (Fig. 6B, compare lane 2 to lanes 4–6). Only at 100 μM chloroquine was a reduction in the level of the Golgi-modified pro-spike observed (lane 3). On the other hand, NH4Cl abrogated the appearance of Golgi-modified forms at ≥10 mM (compare lane 8 with 9–11) and had a milder effect at 1 mM (lane 12). These data clearly demonstrate that the biosynthesis and proteolytic processing of SARS-CoV spike protein are not affected at chloroquine (25 and 50 μM) and NH4Cl (1 mM) doses that cause virus inhibitory effects. In addition, with 40, 20, and 10 mM NH4Cl, there was an increased accumulation of proS-ER with a concomitant decrease in the amount of oligomers (Fig. 6B, lanes 9–11). When we examined the homotrimers, we found that chloroquine at 100 μM and NH4Cl at 40 and 20 mM resulted in slightly faster mobility of the trimers (Fig. 6B, lanes 3, 9, and 10), but lower drug doses, which did exhibit significant antiviral effects, did not result in appreciable differences. These data suggest that the newly synthesized intracellular spike protein may not be a major target for chloroquine and NH4Cl antiviral action. The faster mobility of the trimer at certain higher concentration of the drugs might be due the effect of these drugs on the terminal glycosylation of the trimers.
Discussion
We have identified chloroquine as an effective antiviral agent for SARS-CoV in cell culture conditions, as evidenced by its inhibitory effect when the drug was added prior to infection or after the initiation and establishment of infection. The fact that chloroquine exerts an antiviral effect during pre- and post-infection conditions suggest that it is likely to have both prophylactic and therapeutic advantages. Recently, Keyaerts et al. [21] reported the antiviral properties of chloroquine and identified that the drug affects SARS-CoV replication in cell culture, as evidenced by quantitative RT-PCR. Taken together with the findings of Keyaerts et al. [21], our analysis provides further evidence that chloroquine is effective against SARS-CoV Frankfurt and Urbani strains. We have provided evidence that chloroquine is effective in preventing SARS-CoV infection in cell culture if the drug is added to the cells 24 h prior to infection. In addition, chloroquine was significantly effective even when the drug was added 3–5 h after infection, suggesting an antiviral effect even after the establishment of infection. Since similar results were obtained by NH4Cl treatment of Vero E6 cells, the underlying mechanism(s) of action of these drugs might be similar.
Apart from the probable role of chloroquine on SARS-CoV replication, the mechanisms of action of chloroquine on SARS-CoV are not fully understood. Previous studies have suggested the elevation of pH as a mechanism by which chloroquine reduces the transduction of SARS-CoV pseudotype viruses [17,18]. We examined the effect of chloroquine and NH4Cl on the SARS-CoV spike proteins and on its receptor, ACE2. Immunoprecipitation results of ACE2 clearly demonstrated that effective anti-SARS-CoV concentrations of chloroquine and NH4Cl also impaired the terminal glycosylation of ACE2. However, the flow cytometry data demonstrated that there are no significant differences in the cell surface expression of ACE2 in cells treated with chloroquine or NH4Cl. On the basis of these results, it is reasonable to suggest that the pre-treatment with NH4Cl or chloroquine has possibly resulted in the surface expression of the under-glycosylated ACE2. In the case of chloroquine treatment prior to infection, the impairment of terminal glycosylation of ACE2 may result in reduced binding affinities between ACE2 and SARS-CoV spike protein and negatively influence the initiation of SARS-CoV infection. Since the biosynthesis, processing, Golgi modification, and oligomerization of the newly synthesized spike protein were not appreciably affected by anti-SARS-CoV concentrations of either chloroquine or NH4Cl, we conclude that these events occur in the cell independent of the presence of the drugs. The potential contribution of these drugs in the elevation of endosomal pH and its impact on subsequent virus entry or exit could not be ruled out. A decrease in SARS-CoV pseudotype transduction in the presence of NH4Cl was observed and was attributed to the effect on intracellular pH [17,18]. When chloroquine or NH4Cl are added after infection, these agents can rapidly raise the pH and subvert on-going fusion events between virus and endosomes, thus inhibiting the infection.
In addition, the mechanism of action of NH4Cl and chloroquine might depend on when they were added to the cells. When added after the initiation of infection, these drugs might affect the endosome-mediated fusion, subsequent virus replication, or assembly and release. Previous studies of chloroquine have demonstrated that it has multiple effects on mammalian cells in addition to the elevation of endosomal pH, including the prevention of terminal glycosyaltion of immunoglobulins [22]. When added to virus-infected cells, chloroquine inhibited later stages in vesicular stomatitis virus maturation by inhibiting the glycoprotein expression at the cell surface [23], and it inhibited the production of infectious HIV-1 particles by interfering with terminal glycosylation of the glycoprotein [24,25]. On the basis of these properties, we suggest that the cell surface expression of under-glycosylated ACE2 and its poor affinity to SARS-CoV spike protein may be the primary mechanism by which infection is prevented by drug pretreatment of cells prior to infection. On the other hand, rapid elevation of endosomal pH and abrogation of virus-endosome fusion may be the primary mechanism by which virus infection is prevented under post-treatment conditions. More detailed SARS CoV spike-ACE2 binding assays in the presence or absence of chloroquine will be performed to confirm our findings. Our studies indicate that the impact of NH4Cl and chloroquine on the ACE2 and spike protein profiles are significantly different. NH4Cl exhibits a more pronounced effect than does chloroquine on terminal glycosylation, highlighting the novel intricate differences between chloroquine and ammonium chloride in affecting the protein transport or glycosylation of SARS-CoV spike protein and its receptor, ACE2, despite their well-established similar effects of endosomal pH elevation.
The infectivity of coronaviruses other than SARS-CoV are also affected by chloroquine, as exemplified by the human CoV-229E [15]. The inhibitory effects observed on SARS-CoV infectivity and cell spread occurred in the presence of 1–10 μM chloroquine, which are plasma concentrations achievable during the prophylaxis and treatment of malaria (varying from 1.6–12.5 μM) [26] and hence are well tolerated by patients. It recently was speculated that chloroquine might be effective against SARS and the authors suggested that this compound might block the production of TNFα, IL6, or IFNγ [15]. Our data provide evidence for the possibility of using the well-established drug chloroquine in the clinical management of SARS.
Conclusion
Chloroquine, a relatively safe, effective and cheap drug used for treating many human diseases including malaria, amoebiosis and human immunodeficiency virus is effective in inhibiting the infection and spread of SARS CoV in cell culture. The fact that the drug has significant inhibitory antiviral effect when the susceptible cells were treated either prior to or after infection suggests a possible prophylactic and therapeutic use.
Methods
SARS-CoV infection, immunofluorescence, and immunoprecipitation analyses
Vero E6 cells (an African green monkey kidney cell line) were infected with SARS-CoV (Urbani strain) at a multiplicity of infection of 0.5 for 1 h. The cells were washed with PBS and then incubated in OPTI-MEM (Invitrogen) medium with or without various concentrations of either chloroquine or NH4Cl (both from Sigma). Immunofluorescence staining was performed with SARS-CoV-specific hyperimmune mouse ascitic fluid (HMAF) [8] followed by anti-mouse fluorescein-coupled antibody.
Eighteen hours after infection, the virus-containing supernatants were removed, and the cells were pulsed with 35S-(Cys) for 30 min and chased for 3 h before lysis in RIPA buffer. Clarified cell lysates and media were incubated with HMAF, and immunoprecipitated proteins were separated by 3–8% NuPAGE gel (Invitrogen); proteins were visualized by autoradiography. In some experiments, cells were chased for 3 h with isotope-free medium. Clarified cell supernatants were also immunoprecipitated with SARS-CoV-specific HMAF.
ACE2 flow cytometry analysis and biosynthesis
Vero E6 cells were seeded in Dulbecco's modified Eagle medium (Invitrogen) supplemented with 10% fetal bovine serum. The next day, the cells were incubated in Opti-MEM (Invitrogen) in the presence or absence of 10 μM chloroquine or 20 mM NH4Cl. To analyze the levels of ACE2 at the cell surface, cells were incubated on ice with 10 μg/mL affinity-purified goat anti-ACE2 antibody (R&D Systems) and then incubated with FITC-labeled swine anti-goat IgG antibody (Caltag Laboratories). Labeled cells were analyzed by flow cytometry with a FACSCalibur flow cytometer (BD Biosciences). For ACE2 biosynthesis studies, Vero E6 cells were pulsed with 250 μCi 35S-(Met) (Perkin Elmer) for 3 h with the indicated concentrations of chloroquine or NH4Cl and then lysed in RIPA buffer. Clarified lysates were immunoprecipitated with an affinity-purified goat anti-ACE2 antibody (R&D systems), and the immunoprecipitated proteins were separated by SDS-polyacrylamide gel electrophoresis.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MV did all the experiments pertaining to SARS CoV infection and coordinated the drafting of the manuscript. EB and SB performed experiments on ACE2 biosynthesis and FACS analysis. BE performed data acquisition from the immunofluorescence experiments. PR and TK provided critical reagents and revised the manuscript critically. NS and SN along with MV and EB participated in the planning of the experiments, review and interpretation of data and critical review of the manuscript. All authors read and approved the content of the manuscript.
Acknowledgements
We thank Claudia Chesley and Jonathan Towner for critical reading of the manuscript. This work was supported by a Canadian PENCE grant (T3), CIHR group grant #MGC 64518, and CIHR grant #MGP-44363 (to NGS).
==== Refs
Ksiazek TG Erdman D Goldsmith CS Zaki SR Peret T Emery S Tong S Urbani C Comer JA Lim W Rollin PE Dowell SF Ling AE Humphrey CD Shieh WJ Guarner J Paddock CD Rota PB Fields B DeRisi J Yang JY Cox N Hughes J LeDuc JW Bellini WJ Anderson LJ SARS Working Group A novel coronavirus associated with severe acute respiratory syndrome N Engl J Med 2003 348 1953 1966 12690092 10.1056/NEJMoa030781
Marra MA Jones SJ Astell CR Holt RA Brooks-Wilson A Butterfield YS Khattra J Asano JK Barber SA Chan SY Cloutier A Coughlin SM Freeman D Girn N Griffith OL Leach SR Mayo McDonald H Montgomery SB Pandoh PK Petrescu AS Robertson AG Schein JE Siddiqui A Smailus DE Stott JM Yang GS Plummer F Andonov A Artsob H Bastien N Bernard K Booth TF Bowness D Czub M Drebot M Fernando L Flick R Garbutt M Gray M Grolla A Jones S Feldmann H Meyers A Kabani A Li Y Normand S Stroher U Tipples GA Tyler S Vogrig R Ward D Watson B Brunham RC Krajden M Petric M Skowronski DM Upton C Roper RL The Genome sequence of the SARS-associated coronavirus Science 2003 300 1399 1404 12730501 10.1126/science.1085953
Rota PA Oberste MS Monroe SS Nix WA Campagnoli R Icenogle JP Penaranda S Bankamp B Maher K Chen MH Tong S Tamin A Lowe L Frace M DeRisi JL Chen Q Wang D Erdman DD Peret TC Burns C Ksiazek TG Rollin PE Sanchez A Liffick S Holloway B Limor J McCaustland K Olsen Rasmussen M Fouchier R Gunther S Osterhaus AS Drosten C Pallansch MA Anderson LJ Bellini WJ Characterization of a novel coronavirus associated with severe acute respiratory syndrome Science 2003 300 1394 1399 12730500 10.1126/science.1085952
Ng ML Tan SH See EE Ooi EE Ling AE Proliferative growth of SARS coronavirus in Vero E6 cells J Gen Virol 2003 84 3291 3303 14645910 10.1099/vir.0.19505-0
Li M Moore WJ Vasilieva N Sui J Wong SK Berne MA Somasundaran M Sullivan JL Luzuriaga K Greenough TC Choe H Farzan M Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus Nature 2003 426 450 454 14647384 10.1038/nature02145
Bergeron E Vincent MJ Wickham L Hamelin J Basak A Nichol ST Chrétien M NG Seidah Implication of proprotein convertases in the processing and spread of severe acute respiratory syndrome coronavirus Biochem Biophys Res Comm 2005 326 554 563 15596135 10.1016/j.bbrc.2004.11.063
Zhang Y Li T Fu L Yu C Li Y Xu X Wang Y Ning H Zhang S Chen W Babiuk LA Chang Z Silencing SARS-CoV spike protein expression in cultured cells by RNA interference FEBS Lett 2004 560 141 146 14988013 10.1016/S0014-5793(04)00087-0
Subbarao K McAuliffe J Vogel L Fahle G Fischer S Tatti K Packard M Shieh WJ Zaki S Murphy B Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice J Virol 2004 78 3572 3577 15016880 10.1128/JVI.78.7.3572-3577.2004
Yang ZY Kong WP Huang Y Roberts A Murphy BR Subbarao K Nabel GJ A DNA vaccine induces SARS coronavirus neutralization and protective immunity in mice Nature 2004 428 561 564 15024391 10.1038/nature02463
Bisht H Roberts A Vogel L Bukreyev A Collins PL Murphy BR Subbarao K Moss B Severe acute respiratory syndrome coronavirus spike protein expressed by attenuated vaccinia virus protectively immunizes mice Proc Natl Acad Sci USA 2004 101 6641 6646 15096611 10.1073/pnas.0401939101
Bukreyev A Lamirande EW Buchholz UJ Vogel LN Elkins WR St. Claire M Murphy BR Subbarao K Collins PL Mucosal immunization of African green monkeys (Cercopithecus aethiops) with an attenuated parainfluenza virus expressing the SARS coronavirus spike protein for the prevention of SARS Lancet 2004 363 2122 2127 15220033 10.1016/S0140-6736(04)16501-X
Sainz B JrMossel EC Peters CJ Garry RF Interferon-beta and interferon-gamma synergistically inhibit the replication of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) Virology 2004 329 11 17 15476870 10.1016/j.virol.2004.08.011
Stroher U DiCaro A Li Y Strong JE Aoki F Plummer F Jones SM Feldmann H Severe acute respiratory syndrome-related coronavirus is inhibited by interferon- alpha J Infect Dis 2004 189 1164 1167 15031783 10.1086/382597
Sui J Li W Murakami A Tamin A Matthews LJ Wong SK Moore MJ Tallarico AS Olurinde M Choe H Anderson LJ Bellini WJ Farzan M Marasco WA Potent neutralization of severe acute respiratory syndrome (SARS) coronavirus by a human mAb to S1 protein that blocks receptor association Proc Natl Acad Sci USA 2004 101 2536 2541 14983044 10.1073/pnas.0307140101
Savarino A Boelaert JR Cassone A Majori G Cauda R Effects of chloroquine on viral infections: an old drug against today's diseases? Lancet Infect Dis 2003 3 722 727 14592603 10.1016/S1473-3099(03)00806-5
Ng ML Tan SH See EE Ooi EE Ling AE Early events of SARS coronavirus infection in vero cells J Med Virol 2003 71 323 331 12966536 10.1002/jmv.10499
Simmons G Reeves JD Rennekamp AJ Amberg SM Piefer AJ Bates P Characterization of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) spike glycoprotein-mediated viral entry Proc Natl Acad Sci USA 2004 101 4240 4245 15010527 10.1073/pnas.0306446101
Yang ZY Huang Y Ganesh L Leung K Kong WP Schwartz O Subbarao K Nabel GJ pH-dependent entry of severe acute respiratory syndrome coronavirus is mediated by the spike glycoprotein and enhanced by dendritic cell transfer through DC-SIGN J Virol 2004 78 5642 5650 15140961 10.1128/JVI.78.11.5642-5650.2004
Tipnis SR Hooper NM Hyde R Karran E Christie G Turner AJ A human homolog of angiotensin-converting enzyme. Cloning and functional expression as a captopril-insensitive carboxypeptidase J Biol Chem 2000 275 33238 33243 10924499 10.1074/jbc.M002615200
Song HC Seo MY Stadler K Yoo BJ Choo QL Coates SR Uematsu Y Harada T Greer CE Polo JM Pileri P Eickmann M Rappuoli R Abrignani S Houghton M Han JH Synthesis and characterization of a native, oligomeric form of recombinant severe acute respiratory syndrome coronavirus spike glycoprotein J Virol 2004 78 10328 10335 15367599 10.1128/JVI.78.19.10328-10335.2004
Keyaerts E Vijgen L Maes P Neyts J Ranst MV In vitro inhibition of severe acute respiratory syndrome coronavirus by chloroquine Biochem Biophys Res Commun 2004 323 264 268 15351731 10.1016/j.bbrc.2004.08.085
Thorens B Vassalli P Chloroquine and ammonium chloride prevent terminal glycosylation of immunoglobulins in plasma cells without affecting secretion Nature 1986 321 618 620 3086747 10.1038/321618a0
Dille BJ Johnson TC Inhibition of vesicular stomatitis virus glycoprotein expression by chloroquine J Gen Virol 1982 62 91 103 6290597
Tsai WP Nara PL Kung HF Oroszlan S Inhibition of human immunodeficiency virus infectivity by chloroquine AIDS Res Hum Retroviruses 1990 6 481 489 1692728
Savarino A Lucia MB Rastrelli E Rutella S Golotta C Morra E Tamburrini E Perno CF Boelaert JR Sperber K Cauda RC Anti-HIV effects of chloroquine: inhibition of viral particle glycosylation and synergism with protease inhibitors J Acquir Immune Defic Syndr 2004 35 223 232 15076236
Ducharme J Farinotti R Clinical pharmacokinetics and metabolism of chloroquine. Focus on recent advancements Clin Pharmacokinet 1996 31 257 274 8896943
|
16115318
|
PMC1232869
|
CC BY
|
2021-01-04 16:39:00
|
no
|
Virol J. 2005 Aug 22; 2:69
|
utf-8
|
Virol J
| 2,005 |
10.1186/1743-422X-2-69
|
oa_comm
|
==== Front
Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-801614657110.1186/1743-422X-2-80ResearchNovel type I interferon IL-28A suppresses hepatitis C viral RNA replication Zhu Haizhen [email protected] Mike [email protected] David R [email protected] Chen [email protected] Department of Pathology, Immunology and Laboratory Medicine, University of Florida, P. O. Box 100275, Gainesville, Florida 32610, USA2 Department of Medicine, University of Florida, P. O. Box 100275, Gainesville, Florida 32610, USA2005 7 9 2005 2 80 80 6 6 2005 7 9 2005 Copyright © 2005 Zhu et al; licensee BioMed Central Ltd.2005Zhu 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.
Interferon alpha (IFN-α)-based therapy is the currently approved treatment for chronic hepatitis C viral infection. The sustained antiviral response rate is approximately 50% for genotype-1 infection. The major challenge to the HCV community is to improve antiviral efficacy and to reduce the side effects typically seen in IFNα-based therapy. One of the strategies is to identify new interferons, which may have better efficacy and less undesirable side effects. In this report, we examined the role of IL-28A (IFN λ2), a novel type I IFN, in suppression of human hepatitis C viral RNA replication. We have cloned both the human genomic DNA and cDNA of IL-28A, and evaluated their biological activity using HCV RNA replicon cell culture system. The results show that IL-28A effectively inhibits HCV subgenomic RNA replication in a dose-dependent manner. Treatment of human hepatoma cells with IL-28A activates the JAK-STAT signaling pathway and induces the expression of some interferon-stimulated genes (ISGs), such as 6–16 and 1–8U. We also demonstrate that IL-28A induces expression of HLA class I antigens in human hepatoma cells. Moreover, IL-28A appears to specifically suppress HCV IRES-mediated translation. Although IL-28A receptor shares one subunit with the IL-10 receptor, IL-10 treatment has no detectable effect on IL-28A-induced antiviral activity. Interestingly, IL-28A can synergistically enhance IFNα antiviral efficacy. Our results suggest that IL-28A antiviral activity is associated with the activation of the JAK-STAT signaling pathway and expression of ISGs. The effectiveness of IL-28A antiviral activity and its synergistic effect on IFN-α indicate that IL-28A may be potentially used to treat HCV chronic infection.
==== Body
Background
Interferon alpha (IFN-α), the prototype of type I interferon, is widely used to treat human viral infections and certain malignant tumors [1]. There are several subtypes of type I interferons in humans, namely IFN-α, IFN-β, IFN-ω, IFN-κ, IFN-tau, IFN-epsilon, IFN-zeta, and the recently discovered IFN-λ [2,3]. At least 13 nonallelic IFN-α genes, a single IFN-β gene, and a single IFN-ω gene were identified on human chromosome 9 [4,5]. There are three genes for IFNλ, named as IFN-λ1, IFN-λ2, and IFN-λ3 (also referred to as IL-29, IL-28A, and IL-28B, respectively). Expression of these interferons is induced by viral infection in the majority of nucleated cells. All the type I interferons possess antiviral activity, but the antiviral efficacy appears to vary significantly in subtypes [6,7]. They play a critical role in the innate and adaptive immune responses to viral infection [8]. Interferons exert their biological activities by binding to the heterodimeric receptor. Current evidence suggests that all the type I interferons, except for IFNλ, utilize the same cell membrane-bound receptor, IFNAR, consisting of two subunits, IFNAR1 and IFNAR2. The binding of the receptor by type I interferons predominantly activates The JAK-STAT signaling pathway [9], although other signaling pathways can also be activated in some types of cells [10,11]. Activation of the JAK-STAT pathway leads to induction of the IFN-stimulated gene factor 3 (ISGF), consisting of STAT1, STAT2, and IFN-regulatory factor 9 (IRF-9), which serves as a transcription complex to induce the expression of the downstream target genes, referred to as interferon-stimulated genes (ISG) [12,13]. In either virus-infected or non-infected cells, IFNs induce the transcription of more than 1000 genes [14,15], some of which have been shown to possess direct antiviral properties [16-18]. Moreover, recent studies suggest that type I interferons have an impact on adaptive immunity by regulating MHC class I antigen expression, stimulating dendritic cell maturation [19], and increasing the function of the natural killer (NK) cells [20].
The three members of novel IFNλ have several unique features: 1. The sequence homology of IL-28 and other type I interferons is only 15–19%; 2. These genes contain introns; 3. They bind a specific heterodimeric receptor: one subunit belonging to the class II receptor family and the other subunit is identical to the IL-10 receptor subunit 2; 4. The receptor expression exhibits dramatic variations in different tissues; and, 5. The genes are located on chromosome 19 (q13.13). Despite these unique features of IFN-λ, initial studies have demonstrated that these interferons can be activated by double-stranded RNA and viral infection in cell cultures [2,3]. These interferons suppressed the replication of vesicular stomatitis virus (VSV) and encephalomyocarditis virus (ECMV) in human cell lines, activated the JAK-STAT pathway, and induced expression of some ISGs, which are similar to all the other type I interferons. Thus, it is important to thoroughly investigate these interferons, and to explore the possibility of potential clinical application.
Hepatitis C viral (HCV) infection is a global health problem. It infects more than 170 million people worldwide and 4 million people in the United States [21]. There is no effective vaccine available [22], and the current treatment is the combination therapy with interferon alpha (IFN) and a nucleotide analog, Ribavirin. The best response rate for genotype 1 infection, the predominant viral strain in the United States, is about 50% [23-25]. Moreover, IFN treatment carries significant side effects, partially due to the broad range of IFN biological activities [26]. Unfortunately, the mechanisms of interferon antiviral action, as well as the mechanisms of viral interferon resistance, are still poorly characterized. Thus, a major challenge to the HCV community is to improve therapy for IFN nonresponders, and to reduce its side effects. One of the strategies is to identify new interferons or biological molecules, which may have better efficacy and less undesirable side effects.
In this report, we examined the role of IL-28A in suppression of human HCV RNA replication. We cloned both the human genomic DNA and cDNA of IL-28A, and evaluated their biological activities, cell signaling pathway, and gene induction using HCV RNA replicon cell culture system. We also examined the interactions of IL-28A, IFN, and IL-10.
Results
Cloning of IL-28A genomic DNA and cDNA
To clone the cDNA of IL-28A, we designed primers according to the available sequence information in Genbank (NM_172138). Total RNA was isolated from normal human splenic tissue, followed by RT-PCR amplification. An intense and specific DNA fragment of 1.6 kb in size was identified, which was larger than the predicted 0.7 kb for IL-28A cDNA. DNA sequence analysis confirmed that this fragment represented the genomic DNA of IL-28A gene (Fig. 1). This 1.6 kb PCR reaction product must be derived from the residual DNA in the total RNA preparation. Consistent with this assumption, extensive DNase I treatment of the RNA preparation eliminated the amplification. Because we could not amplify the cDNA fragment using this approach with several attempts, we, therefore, decided to clone the cDNA fragment using the IL-28A genomic clone. First, the IL-28A genomic DNA was cloned into the expression vector pEF/V5-His-TOPO to construct the plasmid, pTOPO-IL-28A. We then transfected pTOPO-IL-28A into Huh-7 cells, followed by RNA isolation and vigorous DNase I digestion. RT-PCR was performed using the same pair of primers described above. A 0.7 kb cDNA fragment was readily obtained, which corresponded to the predicted cDNA size of IL-28A. We sequenced and analyzed both the genomic fragment and its cDNA. As shown in Fig. 1A, there are five introns and six exons in the IL-28A gene. The first ATG starts from nt.53 and ends at nt.655 (Fig 1B). The predicted amino acid sequence was identical to that published by Sheppard et al. [3].
Figure 1 IL-28A gene structure. A) Schematics of the exon-intron structure of the gene. The numbers indicate exon location. B) Complete sequence of IL-28A genomic sequence (Accession number DQ126336). The bold nucleotides are the nucleotide sequence of the exons (Accession number DQ126337).
IL-28A exhibits anti-HCV activity
Since IL-28A is a new member of type I interferon family and IFNα is widely used to treat HCV infection, it is logical to examine its anti-HCV activity. We first tested whether the IL-28A DNA is functional in human hepatoma cells. The eukaryotic expression vectors containing either the genomic DNA or cDNA were transfected into a HCV subgenomic replicon cell line, GSB1. The control cells were transfected by pTOPO vector without any insert sequence. Forty-eight hours after transfection, total RNA was isolated, followed by real-time RT-PCR analysis. As shown in Fig. 2A, compared with the control plasmid-transfected cells, the IL-28A transfected cells had significantly lower viral RNA copy numbers. The viral suppression effect was also demonstrated by viral NS5A protein expression, as determined by Western blot analysis (Fig. 2B). To further determine the effect of IL-28A secreted by cells, we then tested the antiviral effect of the conditioned medium. We subcloned IL-28A cDNA into pEF/V5-His-TOPO vector and generated the plasmid, pTOPO-IL-28A07. The plasmid pTOPO-IL-28A07 was transfected into Huh-7 cells, and the supernatant was harvested after 72 hours of incubation. Varying amounts of the conditioned-medium from Huh7 cells transfected with plasmid pTOPO or pTOPO-IL-28A07 were added to GSB1 cells. After 48 hours of incubation, total RNA was isolated from the cells, followed by real-time RT-PCR analysis. As shown in Fig 2C, the IL-28A-conditioned medium demonstrated a significant inhibitory effect on HCV RNA replication in a dose-dependent manner. Similar results were also obtained using pTOPO-IL28A, the genomic expression construct (data not shown). We then further examined the effect of the recombinant IL-28A on HCV RNA replication by incubating GSB1 cells with varying doses of rhIL-28A, followed by total RNA extraction and real-time PCR analysis. As shown in Fig. 2D, the replication levels of HCV RNA were significantly suppressed by rhIL-28A. Again, IL-28A inhibits the viral RNA replication in a dose-dependent manner, but the effective dose of rhIL28A is significantly higher than recombinant IFN.
Figure 2 Effects of IL-28A on HCV RNA replication and protein expression. A) GSB cells were transfected by either control plasmid (TOPO) or IL-28A genomic expression construct (TOPO-hIL28A). After 48 hours, total RNA was isolated, followed by real-time PCR analysis with HCV-specific primers. The data represents the normalization with the internal control GADPH. B) Western blot analysis of GSB1 cells transfected with the control plasmid (TOPO) or IL-28A expression construct (TOPO-IL28A). The monoclonal antibody is specifically against HCV NS5A. The internal control is actin. C) Effect of IL-28A-conditioned medium on HCV RNA replication in GSB cells. The conditioned medium was used to treat the cells for 48 hours, followed by real-time RT-PCR analysis. D) Effect of recombinant IL-28A on HCV RNA replication in GSB cells. The relative HCV RNA levels were normalized with the internal control GADPH. The error bars indicate the variations of three independent assays.
For simultaneous assessment of cap-dependent and HCV IRES-dependent translation, Huh7 cells were transiently transfected with a bicistronic reporter plasmid, pRL-HL, encoding the Renilla and firefly luciferase cDNAs translated from the 5'cap and internally from the HCV IRES, respectively, for 24 hours, followed by 24 hours of incubation in medium alone or with medium containing increasing amount of hrIL-28A. Cells were harvested, and protein extracts were prepared, and a dual luciferase assay using the luciferase assay system was performed. As shown in Fig. 3, translation from the viral IRES elements exhibited a dose-dependent suppression, while the cap-dependent translation is not significantly affected by IL-28A. These data suggest that IL-28A appears to specifically inhibit HCV IRES-mediated translation without affecting cap-mediated translation in the host cells.
Figure 3 Effects of IL-28A on CAP-dependent and HCV IRES-dependent translation. The GSB1 cells was transfected with control plasmid or plasmid pRL-HL, which has different luciferases directed by either CAP- or HCV IRES. After 48 hours of transfection, the cells were treated with varying doses of IL-28A. Cell extracts were made after 24 hours of incubation, followed by luciferase determination. The data represents the average of three independent experiments. The open column indicates CAP-mediated translation. The shadowed column indicates HCV IRES-mediated translation. The error bars indicate the variations of three independent assays.
IL-28A activates the JAK-STAT signaling pathway
It is known that type I interferons initiate cellular responses at least partially through the JAK-STAT pathway. All the human type I IFNs interact with the same receptor, IFNAR [27]. When IFNs bind to specific cell surface receptors on the host cells, the IFNAR receptor complex will activate the JAK proteins, JAK1 and TYK2. The activated-JAK proteins then phosphorylate STAT1, STAT2, and STAT3. We hypothesized that the IL-28A-induced antiviral effect in GSB1 cells would depend upon the activation of the JAK-STAT signaling pathway. We, therefore, analyzed the status of STAT1 and STAT3 in response to IL-28A stimulation. Huh7 or GSB1 cells were treated with pTOPO-IL-28A07 conditioned medium or rhIL-28A for 30 minutes, followed by total protein extraction and Western blot analysis using anti-p-STAT1, anti-p-STAT3, total STAT1 and STAT3 monoclonal antibodies. As shown in Fig. 4, both phosphorylated STAT1 and STAT3 were detected in cells treated with IL-28A. This result indicates that IL-28A utilizes the similar JAK-STAT signaling pathway as the IFN-α and IFN-β, despite receptor differences.
Figure 4 The effect of IL-28A on JAK-STAT signaling pathway. GSB1 cells were treated with either IL-28A conditioned medium (A) or recombinant rIL-28A (B) for 30 minutes, followed by protein extraction and Western blot analysis using antibodies as indicated in the figure. Equal amounts (20 ug) of proteins were loaded in each lane and confirmed by detection of actin. STAT1 or STAT3 indicates total STAT protein. p-STAT1 or p-STAT3 indicates tyrosine phosphorylated form (activated STAT protein). The figures are representatives of at least four independent experiments.
IL-28A induces interferon stimulated genes (ISGs) expression
The transcription factor IFN-stimulated gene factor 3 complex (ISGF3), consisting of phosphorylated STAT1, phosphorylated STAT2, and IRF-9/p48, translocates into the nucleus and binds to IFN-stimulated response elements (ISRE) within the promoters of ISGs [9]. Interferons exert their biological function through induction of ISGs in the cell. Therefore, it is possible that IL-28A provides antiviral activity by induction of a subset of IFN-stimulated genes (ISGs). To determine whether IL-28A can induce the ISGs, total RNA was isolated from the cells treated by IL-28A-conditioned medium from Huh7 cells transfected with pTOPO-IL-28A, followed by semi-quantitative RT-PCR analysis using gene specific primer sets for 6–16, 1–8U, 1–8D, and IFIT1. As shown in Fig. 5, 6–16 and 1–8U were significantly induced by IL-28A, while the gene IFIT1 was not effectively induced. This observation suggests that IL-28A is capable of inducing ISGs, but the gene profile may not be identical to that of IFN.
Figure 5 Induction of interferon stimulated genes by IL-28A in GSB1 cells. GSB1 cells were treated by either control or 2-ml IL-28A-conditioned medium for 12 hours. Total RNA was isolated, followed by RT-PCR analysis using a pair of gene-specific primers and a pair of DADPH primers. The PCR amplification cycle is 25, which ensures PCR reaction in linear range. The PCR products were analyzed in 1% agarose gel. M indicates the DNA molecular weight marker. The arrow indicates gene-specific products. The bar indicates GADPH DNA fragment. The figure is a representative of two independent assays.
IL-10 has no effect on the IL-28A-induced anti-HCV activity
The IL-28A receptor complex consists of a ligand-binding chain, IL-28R, and an accessory receptor chain, IL-10R2. So it is logical to determine whether IL-10 interferes with IL-28A in inhibiting HCV RNA replication. GSB1 cells were treated with or without IL-28A (100 ng/mL and 300 ng/mL) in the presence or absence of 100 ng/mL IL-10. After 72 hours of incubation, total RNA was isolated from the cells, followed by real-time RT-PCR analysis. As shown in Fig. 6, IL-10 did not have a significant effect on the IL-28A-induced anti-HCV activity, while IL-28A can decrease RNA replication,
Figure 6 Effect of IL-10 on IL-28A-induced antiviral activity. The GSB1 cells were treated with varying doses of IL-10 or IL-28A as indicated for 72 hours. Total RNA was isolated for real-time PCR analysis using HCV-specific primers. The data represents the normalization with internal control GADPH. The error bars indicate the variations of three independent assays.
IL-28A synergies with IFN-α in suppressing HCV RNA replication
The above results indicated that IL-28A can signal through the JAK-STAT pathway in a similar manner as to IFN-α. To determine whether IL-28A enhances IFN-α-induced anti-HCV RNA replication, we tested the effect of IL-28A on IFN-α-induced anti-HCV RNA activity using GSB1 cell. IFN-α was used at the dose of 50 U/mL, 100 U/mL with or without 100 ng/mL IL-28A. After 24 hours of incubation, cells were harvested and total RNA was isolated, followed by real-time RT-PCR analysis using HCV-specific primers. As shown in Fig. 7, the combination of IL-28A and IFN-α reduced HCV viral RNA by 100-fold, while IFNα alone reduced the virus by 15-fold and IL-28A alone by 6-fold. Activation of STAT1 protein by phosphorylation is a critical step for IFN-α signaling pathway. In the next experiment, we examined the effect of IL-28A on the IFN-α-induced STAT1 phosphorylation by Western blotting. As shown in Fig. 8, the levels of p-STAT1 were significantly higher than those induced by IFN-α or IL-28A alone. In addition, STAT1 remained phosphorylated for 8 h after stimulation with IFN-α plus IL-28A, while STAT1 phosphorylation induced by IFN treatment alone decreased to undetectable level (data not shown). The results indicate that IL-28A can synergize with IFN-α in suppressing the HCV RNA replication and inducing intracellular antiviral signaling pathway.
Figure 7 Effect of IL-28A on the antiviral activity of IFNα. Varying doses of IL-28A and IFNα2b, either alone or in combination, were added to the GSB1 cells and incubated for 48 hours. Total RNA was isolated for real-time PCR analysis. The vertical axis represents the fold of viral RNA reduction by IL-28A or IFN. The data represents the results of normalization with the internal control GADPH.
Figure 8 Effect of IL-28A on IFNα-induced STAT1 activation in GSB1 cells. GSB1 cells were treated with IL-28A or/and IFN as indicated. After 30 minutes of incubation, total protein was extracted for Western blot analysis using antibodies against total STAT1 (STAT1) or phosphorylation-specific STAT1 (p-STAT1).
IL-28A induces HLA class I antigen expression
Type I interferons are believed to play a role in immune regulation. One of the mechanisms is through induction of HLA class I antigen. To test whether IL-28A has such an effect, we treated Huh7 cells with IL-28A-conditioned medium from Huh7 cells transfected by plasmid pTOPO-IL-28A, followed by flow cytometric analysis using anti-HLA class I antigen. As shown in Fig. 9, treatment with IL-28A induced HLA class I antigen production. The data suggest that IL-28A has a similar capacity to induce class I antigen production as other type I IFN.
Figure 9 Effect of IL-28A on HLA class I antigen expression in Huh7 cells. The Huh7 cells were treated with 2 ml IL-28A-conditioned culture medium for 72 hours. The cells were then harvested and incubated with HLA class I antigen-specific antibody labeled by FITC fluorescence, followed by flow cytometric analysis. The arrow-marked curve indicates control cells. The arrowhead-marked curve indicates cells treated with IL-28A.
Discussion
Type I interferons play an essential role in innate immune responses against viral infections. There are many subtypes of type I interferons in humans, including the recently identified IFN-λ, consisting of three members, λ1 (IL-29), λ2 (IL-28A), and λ3 (IL-28B). The most extensively studied subtypes are IFN-α and IFN-β. There is relatively little information available for IFN-λ. The major difference between IFN-λ and the other type I IFNs is the utilization of different receptors. Current type I interferon therapy has significant side effects. Identification of novel type I interferons with desirable clinical efficacy and less side effects is needed. IFNλ s are potentially such candidates.
In this report, we have cloned both the cDNA and the gene of human IL-28A. Through a series experiments, we have shown the biological effects of this protein on HCV viral replication, its signaling events in human liver cells, and its interaction with IFN-α and IL-10.
To clone this gene, we employed a RT-PCR approach using total RNA extracted from spleen, liver, and peripheral blood mononuclear cells (PBMC). With extensive effort, we could only obtain IL-28A genomic clones but not cDNA, while we could readily amplify IFN-α and IFN-β cDNA from the same RNA source. We confirmed that the amplified genomic clones were derived from the residual DNA in the RNA preparation, since two rounds of DNase I treatment eliminated the amplification. This indicates that there is no detectable IL-28A expression in these tissues at a normal physiological condition, although it has been reported that IL-28A is expressed in PBMCs from HCV-infected patients [28]. To obtain the cDNA clone, we decided to clone the genomic DNA into a expression vector, and then transfected it into Huh7 cells. Total RNA was extracted from the transfected cells and RT-PCR was performed. The cDNA DNA fragment was easily amplified using this approach. We noticed that Kotenko et al. used a similar strategy to clone the first IL-28A cDNA [2]. By comparing the cDNA and its gene, we identified five introns and six extrons. So far, this is the only type I interferon gene containing introns, while the other type I IFNs encode within a single extron. The presence of multiple introns makes this gene more similar to IL-10 gene family. Interestingly, the IL-28A receptor shares one subunit with IL10 (IL-10Rβ). We know that IL-10 and type I interferons play a different role during the host immune responses to viral infections. The presence of introns generally subjects the gene to an additional gene expression control. According to Kotenco et al., the IL-28A is predominantly expressed in the heart, liver and spleen [2]. Whether the introns play any role in such relatively tissue-restricted expression remains to be investigated.
After cloning this gene, we then showed that the gene product, IL-28A, has similar biological properties as other type I interferons. IL-28A resembles type I IFNs in its ability to induce anti-HCV activity through JAK-STAT signaling pathway. As we have shown in Fig. 4, IL-28A activates both STAT1 and STAT3. The IL-28A-mediated antiviral activity is dose-dependent. Both the recombinant and the gene product produced in liver cells are effective, though the effective dose of the recombinant IL-28A is much higher than the other type I interferons. Similar results were recently reported by other laboratories [29,30].
We further analyzed the expression of ISGs using a RT-PCR approach. Interestingly, at least one ISG cannot be induced by IL-28A, while it can be readily induced by IFNα. Moreover, by testing the effect of interferons on cap-mediated translation and HCV IRES-mediated translation, our preliminary data showed that IL-28A appears to have a selective activity to inhibit HCV-IRES-mediated translation, while it did not affect cap-mediated translation. This observation is consistent with the fact: even at higher dose (1000 ng/ml), IL-28A did not exhibit antiproliferation activity in a human hepatoma cell line (data not shown). These data suggest that IL-28A seems to have at least some different biological activities as compared with IFN-α. Whether these differences can be employed to achieve therapeutic advantage remains to be determined.
As we have mentioned above, the receptor for IL-28A shares a common subunit with IL-10. Our previous study showed that IL-10 did not have direct antiviral activity in patients with chronic HCV infection [31]. We asked the question whether the sharing of a receptor has any impact on IL-28A activity. Our data suggests that IL-10 does not have an antiviral effect in HCV replicon cells, nor does it have any interference with IL-28A antiviral effect. Thus, the significance of receptor sharing remains unknown.
Since IL-28A and other type I interferon use different receptor for signaling transduction, we next examined the combination effect of IL-28A and IFN-α. Interestingly, combination of IL-28A and IFN-α exhibited synergistic effect on JAK-STAT activation and anti-HCV activity. As shown in Figure 7, combination of 50 U IFN-α and 100 ng per milliliter IL-28A reduced HCV RNA by 40 folds, while individual IFN-α and IL-28A reduced HCV RNA by 10-fold and 6-fold, respectively. We do not know the precise mechanism of this synergistic effect, though the STAT1 activation shows the similar synergistic effect (Fig. 8). It is possible that the activation of one receptor may have beneficial effect on the other receptor-mediated pathway. It is also possible that the common downstream molecules shared by both pathways can synergistically induced by these two interferons. This synergistic effect has a significant clinical implication. It is tempting to speculate that combination of these two reagents may have therapeutic benefit for HCV therapy, particularly in the setting of IFN resistance.
Type I interferons have an immunoregulatory function [32,33]. One of the mechanisms is through induction of HLA class I antigens [34]. We tested whether IL-28A has a similar activity. Human hepatoma cells have relatively lower HLA class I antigen expression comparing with normal hepatocytes [35]. Treatment of the hepatoma cells increased class I antigen expression through flow cytometric study. Not only this shows that the IL-28A has immunoregulator effect, but the fact that IL-28A can induce HLA class I antigen in tumor cells may implicate the role of IL-28A in tumor immune therapy. It would be interesting to see whether IL-28A is capable of promoting the host antitumor immunity.
Conclusion
Our study shows the gene structure of IL-28A, its antiviral effect on HCV, its signaling transduction pathway, and the induction of ISGs. More importantly, we demonstrate the synergistic effect of IL-28A and IFNα on anti-HCV activity, which has a potential clinical application. IFN-α is currently used for the treatment for chronic HCV infection, HBV infection, and many malignant tumors, including hepatitis B, melanoma, hairy cell leukemia, and non-Hodgkin's lymphoma. IL-28A is a potential therapeutic agent to treat these clinical diseases.
Methods
Cell cultures, reagents and plasmids
The HCV subgenomic replicon cell line, GSB1, was a gift from Dr. Christopher Seeger [36,37]. All cells were propagated in DMEM supplemented with 10% FBS, 200 μM L-glutamine, nonessential amino acids, penicillin and streptomycin. Culture of the replicon cells has been previously described [15]. The expression vector, pEF6/V5-His-TOPO, was obtained from Invitrogen (Carlsbad, CA). The HCV-NS5A-specific monoclonal antibody was generated in the laboratory. Monoclonal antibodies against actin, STAT1, STAT3 and phosphorylated STAT3 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The antibodies against phosphorylated STAT1 were obtained from Upstate (Charlottesville, VA). The secondary antibody goat anti-mouse or anti-rabbit IgG-HRP was from Santa Cruz Biotechnology. Supersignal West Pico Chemiluminescent Substrate was purchased from Pierce Biotechnology, Inc. (Rockford, IL). Recombinant human IL-28A (rhIL-28A) and hIL-10 were purchased from R&D Systems (Mineanapolis, MN). The plasmid pRL-HL (a gift from Dr. Lemon) is a bicistronic expression construct encoding Renilla and firefly luciferase cDNAs translated from 5'cap and internally from the HCV IRES (internal ribosome entry site), respectively [38].
Amplification of human IL-28A DNA, cDNA, and plasmid construction
RNA was isolated from human spleen. The human IL-28A cDNA was amplified by RT-PCR from human spleen RNA using two primers: 5'-GGGTGACAGCCTCAGAGTG-3', 5'-ATAGCGACTGGGTGGCAATA-3'. Superscript One-Step RT-PCR kit with platinum Taq according to the instructions (Invitrogen). The One-Step RT-PCR conditions were as follows: 50°C, 30 min; 94°C, 4 min; followed with 40 cycles (95°C, 30 s; 55°C, 30 s; 72°C, 1 min;). The IL-28A DNA was ligated into pEF6/V5-His-TOPO vector. The expression vector pTOPO-IL-28A were transfected into Huh7 cells using Lipofectin Reagent (Invitrogen) according to the manufacturer's instruction. The total RNA was purified from Huh7 cells transfected by pTOPO-IL-28A for 48 hours and treated by DNase I. The human IL-28A cDNA was generated by RT-PCR from the total RNA pretreated by DNase using the above primers. The reactions were performed using 72°C, 7 mins. The expression vector pTOPO-hIL-28A 0.7 was constructed by inserting the human IL-28A cDNA into pEF6/V5-His-TOPO.
Human IL-28A DNA Sequencing
The IL-28A DNA was amplified as described above. The expression vector TOPO-hIL-28A was sequenced using The BigDye Terminator V3.1 Kit from Applied Biosystems (Foster City, CA). The reaction condition was: 96°C, 10 s; 50°C, 5 s; 60°C, 4 min, total 25 cycles. After that, 1/20 volume of 3 M sodium acetate (pH5.2) and 3 times volume of ethanol were added, and incubated at -20°C for 30 mins, followed by spinning down at 13000 g at 4°C for 30 mins. The DNA pellet was washed using 70% ethanol and dried by vacuum. The sequence was detected by ABI PRISM 377 DNA Sequencer (Applied Biosystems).
DNA transfection
The transfection protocol has been described previously [39,40]. Briefly, GSB or Huh7 cells were transfected with control plasmid pTOPO, pTOPO-IL-28A or pTOPO-IL-28A07 plasmid using Lipofectin. In a 6-well tissue culture plate, 1 × 105 GSB or Huh7 cells were seeded in 2 ml of DMEM supplemented with serum and incubate at 37°C in an incubator overnight. For each transfection, 2 μg of DNA was used. The plasmid, pTOPO, pTOPO-IL-28A, or pTOPO-IL-28A07 was transiently transfected into GSB or Huh7 cells. The transfected cells were incubated for another 48 hours before experiments.
Reverse Transcription and Polymerase Chain Reaction (RT-PCR)
Total cellular RNA was purified from cells. After reverse transcription, cDNA was used for PCR. The primers are for 6–16 (G1P3), forward 5'-AACCGTTTACTCGCTGCTGT-3, reverse 5'-GCTGCTGGCTACTCCTCA-3'; for 1–8U, forward 5'-CAAATGCCAGGAAAAGGAA-3', reverse 5'-ATACAGGTCATGGGCAGAGC; for 1–8D, forward 5'-TGCCAGGAA GAGGAAACTGT-3', reverse 5'-CCTCAATGATGCCTCCTGAT-3'; for IFIT1, forward 5'-TCTCAGAGGAGCCTGGCTAA-3', reverse 5'-AGTGGCTGATATCT GGGTGC-3'; for GAPDH, forward 5-TCACCAGGGCTGCTTTTA-3', reverse 5'-TTCACACCCATGACGAACA-3'. The PCR conditions were as follows: 94°C, 4 min; (95°C, 30 s; 55°C, 30 s; 72°C, 1 min;) × 40 cycles; 72°C, 7 mins. The PCR product was detected on 2% agarose gel.
Quantitative Real-Time PCR
Total cellular RNA was isolated from cells as described before. Real-time PCR was preformed as described previously [39]. Briefly, first-strand cDNAs were synthesized from total cellular RNA by reverse transcription (20 μl of reaction volume) using the Superscript II (50 U reverse transcriptase per reaction) first-strand synthesis for RT-PCR kit (Invitrogen) primed with oligo (dT)12–18 (Invitrogen) according to the manufacturer's instructions. Fluorophore-labeled LUX primers and their unlabeled counterparts were obtained from Invitrogen. Reactions were conducted in a 96-well spectrofluorometric thermal cycler (ABI PRISM 7700 Sequence detector system, Applied Biosystems). Fluorescence was monitored during every PCR cycle at the annealing step. The primers for HCV are: 5'-CGCTCAATGCCTGGAGATTTG-3', 5'-GCACTCGCAAGCACCCTATC-3'; for GADPH: 5'-TGCTGGCGCTGAGTACGTC-3', 5'-GTGCAGGAGGCATTGCTGA-3'. PCR conditions were as follows: 50°C, 2 min; 95°C, 10 min; (95°C, 15 s; 60°C, 1 min) × 40 cycles. Results were analyzed with SDS 2.0 software from Applied Biosystems. Results for all experiments represent triplicate determinations. Results are represented as means ± SD.
Western Blot Analysis
Equal numbers of cells were washed with PBS and lysed in RIPA buffer as described previously [15]. Protein extraction from cells, electrophoresis and Western blot analysis were described previously. Approximately 20 μg of protein were electrophoresed on a 8% SDS-polyacrylamide gel and transferred to polyvinylidene difluoride membrane (Bio-Rad). The membrane was incubated overnight at 4°C in a block buffer (TBS containing 0.1% Tween 20 and 5% fat-free milk power). The blots were probed with monoclonal antibodies specific for NS5A, STAT1, and STAT3, p-STAT3, actin or polyclonal antibody specific for p-STAT1 for 1 hour at room temperature. After being washed 3 times for 30 min each with 0.1% Tween 20 in TBS, the membrane was incubated with the secondary antibody diluted in 5% fat-free milk in TBS containing 0.1% Tween 20 for 1 hour at room temperature and washed 3 times as described above. Proteins were visualized by using Supersignal West Pico Chemiluminescent Substrate.
Flow cytometry
To detect the expression of MHC class I antigen, Huh7 cells were treated with IL-28A conditioned medium from Huh7 cells transfected by plasmid pTOPO-IL-28A for 72 hours and their MHC class I expression was analyzed by flow cytometry as previously described [41]. Cell surface expression of the HLA class I antigens were detected using class I antibody, followed by fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG. Ligand binding was detected by flow cytometry.
Acknowledgements
We thank Drs. James Crawford, Jinxiong She, John Elyer, and Christopher Seeger for the helpful discussion. The work was supported in part by the Charles Trey MD Memorial liver scholar award from American Liver Foundation and DK02958 from NIH to C.L.
==== Refs
Brassard DL Grace MJ Bordens RW Interferon-alpha as an immunotherapeutic protein J Leukoc Biol 2002 71 565 581 11927642
Kotenko SV Gallagher G Baurin VV Lewis-Antes A Shen M Shah NK Langer JA Sheikh F Dickensheets H Donnelly RP IFN-lambdas mediate antiviral protection through a distinct class II cytokine receptor complex Nat Immunol 2003 4 69 77 12483210 10.1038/ni875
Sheppard P Kindsvogel W Xu W Henderson K Schlutsmeyer S Whitmore TE Kuestner R Garrigues U Birks C Roraback J Ostrander C Dong D Shin J Presnell S Fox B Haldeman B Cooper E Taft D Gilbert T Grant FJ Tackett M Krivan W McKnight G Clegg C Foster D Klucher KM IL-28, IL-29 and their class II cytokine receptor IL-28R Nat Immunol 2003 4 63 68 12469119 10.1038/ni873
Chen J Baig E Fish EN Diversity and relatedness among the type I interferons J Interferon Cytokine Res 2004 24 687 698 15684736
Gray PW Goeddel DV Structure of the human immune interferon gene Nature 1982 298 859 863 6180322 10.1038/298859a0
Stark GR Kerr IM Williams BR Silverman RH Schreiber RD How cells respond to interferons Annu Rev Biochem 1998 67 227 264 9759489 10.1146/annurev.biochem.67.1.227
Foster GR Finter NB Are all type I human interferons equivalent? J Viral Hepat 1998 5 143 152 9658366 10.1046/j.1365-2893.1998.00103.x
Samuel CE Antiviral actions of interferons Clin Microbiol Rev 2001 14 778 809, table of contents 11585785 10.1128/CMR.14.4.778-809.2001
Darnell JEJ Kerr IM Stark GR Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins Science 1994 264 1415 1421 8197455
Larner A Reich NC Interferon signal transduction Biotherapy 1996 8 175 181 8813329 10.1007/BF01877202
Parmar S Platanias LC Interferons: mechanisms of action and clinical applications Curr Opin Oncol 2003 15 431 439 14624225 10.1097/00001622-200311000-00005
Williams BR Haque SJ Interacting pathways of interferon signaling Semin Oncol 1997 24 S9 70-S9-77 9208875
Liu KD Gaffen SL Goldsmith MA JAK/STAT signaling by cytokine receptors Curr Opin Immunol 1998 10 271 278 9638363 10.1016/S0952-7915(98)80165-9
Der SD Zhou A Williams BR Silverman RH Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays Proc Natl Acad Sci U S A 1998 95 15623 15628 9861020 10.1073/pnas.95.26.15623
Zhu H Zhao H Collins CD Eckenrode SE Run Q McIndoe RA Crawford JM Nelson DR She JX Liu C Gene expression associated with interferon alfa antiviral activity in an HCV replicon cell line Hepatology 2003 37 1180 1188 12717400 10.1053/jhep.2003.50184
Kochs G Haller O Interferon-induced human MxA GTPase blocks nuclear import of Thogoto virus nucleocapsids Proc Natl Acad Sci U S A 1999 96 2082 2086 10051598 10.1073/pnas.96.5.2082
Sen GC Viruses and interferons Annu Rev Microbiol 2001 55 255 281 11544356 10.1146/annurev.micro.55.1.255
Durbin RK Mertz SE Koromilas AE Durbin JE PKR protection against intranasal vesicular stomatitis virus infection is mouse strain dependent Viral Immunol 2002 15 41 51 11952146 10.1089/088282402317340224
Buelens C Bartholome EJ Amraoui Z Boutriaux M Salmon I Thielemans K Willems F Goldman M Interleukin-3 and interferon beta cooperate to induce differentiation of monocytes into dendritic cells with potent helper T-cell stimulatory properties Blood 2002 99 993 998 11807004 10.1182/blood.V99.3.993
Marrack P Kappler J Mitchell T Type I interferons keep activated T cells alive J Exp Med 1999 189 521 530 9927514 10.1084/jem.189.3.521
Alter MJ The epidemiology of acute and chronic hepatitis C Clin Liver Dis 1997 1 559 68, vi-vii 15560058 10.1016/S1089-3261(05)70321-4
Lechmann M Liang TJ Vaccine development for hepatitis C Semin Liver Dis 2000 20 211 226 10946425 10.1055/s-2000-9947
Fried MW Shiffman ML Reddy KR Smith C Marinos G Goncales FLJ Haussinger D Diago M Carosi G Dhumeaux D Craxi A Lin A Hoffman J Yu J Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection N Engl J Med 2002 347 975 982 12324553 10.1056/NEJMoa020047
Manns MP McHutchison JG Gordon SC Rustgi VK Shiffman M Reindollar R Goodman ZD Koury K Ling M Albrecht JK Peginterferon alfa-2b plus ribavirin compared with interferon alfa-2b plus ribavirin for initial treatment of chronic hepatitis C: a randomised trial Lancet 2001 358 958 965 11583749 10.1016/S0140-6736(01)06102-5
Pawlotsky JM The nature of interferon-alpha resistance in hepatitis C virus infection Curr Opin Infect Dis 2003 16 587 592 14624110 10.1097/00001432-200312000-00012
Matthews SJ McCoy C Peginterferon alfa-2a: a review of approved and investigational uses Clin Ther 2004 26 991 1025 15336466 10.1016/S0149-2918(04)90173-7
Pestka S Langer JA Zoon KC Samuel CE Interferons and their actions Annu Rev Biochem 1987 56 727 777 2441659 10.1146/annurev.bi.56.070187.003455
Mihm S Frese M Meier V Wietzke-Braun P Scharf JG Bartenschlager R Ramadori G Interferon type I gene expression in chronic hepatitis C Lab Invest 2004 84 1148 1159 15208644 10.1038/labinvest.3700135
Robek MD Boyd BS Chisari FV Lambda interferon inhibits hepatitis B and C virus replication J Virol 2005 79 3851 3854 15731279 10.1128/JVI.79.6.3851-3854.2005
Brand S Zitzmann K Dambacher J Beigel F Olszak T Vlotides G Eichhorst ST Goke B Diepolder H Auernhammer CJ SOCS-1 inhibits expression of the antiviral proteins 2',5'-OAS and MxA induced by the novel interferon-lambdas IL-28A and IL-29 Biochem Biophys Res Commun 2005 331 543 548 15850793 10.1016/j.bbrc.2005.04.004
Nelson DR Tu Z Soldevila-Pico C Abdelmalek M Zhu H Xu YL Cabrera R Liu C Davis GL Long-term interleukin 10 therapy in chronic hepatitis C patients has a proviral and anti-inflammatory effect Hepatology 2003 38 859 868 14512873 10.1053/jhep.2003.50427
Biron CA Interferons alpha and beta as immune regulators--a new look Immunity 2001 14 661 664 11420036 10.1016/S1074-7613(01)00154-6
Tompkins WA Immunomodulation and therapeutic effects of the oral use of interferon-alpha: mechanism of action J Interferon Cytokine Res 1999 19 817 828 10476925 10.1089/107999099313325
Girdlestone J Regulation of HLA class I loci by interferons Immunobiology 1995 193 229 237 8530148
Kurokohchi K Carrington M Mann DL Simonis TB Alexander-Miller MA Feinstone SM Akatsuka T Berzofsky JA Expression of HLA class I molecules and the transporter associated with antigen processing in hepatocellular carcinoma Hepatology 1996 23 1181 1188 8621152
Guo JT Bichko VV Seeger C Effect of alpha interferon on the hepatitis C virus replicon J Virol 2001 75 8516 8523 11507197 10.1128/JVI.75.18.8516-8523.2001
Zhu Q Guo JT Seeger C Replication of hepatitis C virus subgenomes in nonhepatic epithelial and mouse hepatoma cells J Virol 2003 77 9204 9210 12915536 10.1128/JVI.77.17.9204-9210.2003
Honda M Kaneko S Shimazaki T Matsushita E Kobayashi K Ping LH Zhang HC Lemon SM Hepatitis C virus core protein induces apoptosis and impairs cell-cycle regulation in stably transformed Chinese hamster ovary cells Hepatology 2000 31 1351 1359 10827163 10.1053/jhep.2000.7985
Zhu H Shang X Terada N Liu C STAT3 induces anti-hepatitis C viral activity in liver cells Biochem Biophys Res Commun 2004 324 518 528 15474458 10.1016/j.bbrc.2004.09.081
Zhu H Liu C Interleukin-1 inhibits hepatitis C virus subgenomic RNA replication by activation of extracellular regulated kinase pathway J Virol 2003 77 5493 5498 12692250 10.1128/JVI.77.9.5493-5498.2003
Shang XZ Zhu H Lin K Tu Z Chen J Nelson DR Liu C Stabilized beta-catenin promotes hepatocyte proliferation and inhibits TNFalpha-induced apoptosis Lab Invest 2004 84 332 341 14767485 10.1038/labinvest.3700043
|
16146571
|
PMC1232870
|
CC BY
|
2021-01-04 16:38:58
|
no
|
Virol J. 2005 Sep 7; 2:80
|
utf-8
|
Virol J
| 2,005 |
10.1186/1743-422X-2-80
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1617612110.1371/journal.pbio.0030334Research ArticleGenetics/Genomics/Gene TherapyMicrobiologyBiochemistryEubacteriaTwo-Component Signal Transduction Pathways Regulating Growth and Cell Cycle Progression in a Bacterium: A System-Level Analysis Systematic Analysis of Two-Component SignalingSkerker Jeffrey M
1
Prasol Melanie S
1
¤Perchuk Barrett S
1
Biondi Emanuele G
1
Laub Michael T [email protected]
1
1Bauer Center for Genomics Research, Harvard University, Cambridge, Massachusetts, United States of AmericaArkin Adam Academic EditorLawrence Berkeley National LaboratoryUnited States of America10 2005 27 9 2005 27 9 2005 3 10 e33426 5 2005 22 7 2005 Copyright: © 2005 Skerker 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.
Mapping Core Communication Networks in Bacteria
Two-component signal transduction systems, comprised of histidine kinases and their response regulator substrates, are the predominant means by which bacteria sense and respond to extracellular signals. These systems allow cells to adapt to prevailing conditions by modifying cellular physiology, including initiating programs of gene expression, catalyzing reactions, or modifying protein–protein interactions. These signaling pathways have also been demonstrated to play a role in coordinating bacterial cell cycle progression and development. Here we report a system-level investigation of two-component pathways in the model organism Caulobacter crescentus. First, by a comprehensive deletion analysis we show that at least 39 of the 106 two-component genes are required for cell cycle progression, growth, or morphogenesis. These include nine genes essential for growth or viability of the organism. We then use a systematic biochemical approach, called phosphotransfer profiling, to map the connectivity of histidine kinases and response regulators. Combining these genetic and biochemical approaches, we identify a new, highly conserved essential signaling pathway from the histidine kinase CenK to the response regulator CenR, which plays a critical role in controlling cell envelope biogenesis and structure. Depletion of either cenK or cenR leads to an unusual, severe blebbing of cell envelope material, whereas constitutive activation of the pathway compromises cell envelope integrity, resulting in cell lysis and death. We propose that the CenK–CenR pathway may be a suitable target for new antibiotic development, given previous successes in targeting the bacterial cell wall. Finally, the ability of our in vitro phosphotransfer profiling method to identify signaling pathways that operate in vivo takes advantage of an observation that histidine kinases are endowed with a global kinetic preference for their cognate response regulators. We propose that this system-wide selectivity insulates two-component pathways from one another, preventing unwanted cross-talk.
Histidine kinases and their (sensory) response regulators are screened for in C. crescentus. Follow-up experiments determine several essential components, including one pair critical for cell envelope biogenesis and structure.
==== Body
Introduction
Cells have the remarkable ability to sense, respond to, and adapt to their internal and external environments in order to maximize survival or accurately execute a developmental program. Such behavior requires the ability to process information, and cells have evolved complex regulatory and signaling systems capable of sophisticated information-processing tasks. It is ultimately the wiring of such systems and the relative quantitative strength of connections that confer on cells the ability to make decisions and regulate their behavior. Thus, there is a need to develop comprehensive, genome-wide maps of the complex signaling pathways operating inside cells. Although transcriptional networks in many organisms have recently been mapped on a global level using DNA microarrays, signaling pathways and networks can be considerably more difficult to study in a systematic, comprehensive fashion, requiring experimentally tractable systems amenable to a combination of genetic and biochemical methods.
Here we report the design and use of a suite of tools for the rapid, systematic mapping of signaling networks responsible for regulating growth, cell cycle progression, and differentiation in the Gram-negative bacterium Caulobacter crescentus. This organism has emerged as an excellent model system for studying regulation of cell cycle progression and development owing to its dimorphic lifestyle (Figure 1A) [1–3]. Each cell division produces two different daughter cells: a stalked cell and a swarmer cell. The motile, chemotactic swarmer cell is unable to initiate DNA replication. In response to poorly understood environmental and internal cues, a swarmer cell differentiates into a stalked cell by losing its polar flagellum, chemotaxis machinery, and polar pili, followed by growth of a stalk. This motile-to-sessile transition is accompanied by increased rates of growth and protein synthesis [4]. This transition also coincides with DNA replication initiation and is thus a G1–S cell cycle transition. A single round of DNA replication ensues, followed by segregation of the daughter chromosomes to opposite ends of the predivisional cell. The development of the predivisional cell includes construction of a new flagellum, chemotaxis machinery, and pili secretion apparatus at the pole opposite the stalk. Cell division is asymmetric, producing a stalked cell that immediately reinitiates another round of DNA replication and a swarmer cell that will again differentiate into a stalked cell. Swarmer cells can be easily isolated from a mixed population of cells by density centrifugation and followed as they proceed synchronously through the cell cycle.
Figure 1 The C. crescentus Cell Cycle and Two-Component Signal Transduction
(A) Schematic diagram of progression through the C. crescentus cell cycle, as described in the text. The timing of key cell cycle and developmental events are indicated. Cell division is asymmetric, generating two distinct daughter cells. The stalked cell can immediately initiate DNA replication, whereas the swarmer cell must first differentiate into a stalked cell.
(B) Diagram of a canonical two-component signal transduction system. On receipt of an input signal, the histidine kinase autophosphorylates on a conserved histidine residue. The phosphoryl group is then passed to the receiver domain of a cognate response regulator. Phosphorylation of the receiver domain typically activates the output domain, which can execute a variety of cellular tasks including initiating programs of gene expression, catalyzing metabolic reactions, or modifying protein–protein interactions.
The regulation of this complex life cycle centers on a single class of signaling molecules known as two-component signal transduction systems. These systems are one of the key signaling modalities in the bacterial kingdom, as well as being present in fungi, slime molds, and plants [5]. As they appear to be absent from metazoans, including humans, this class of molecules has been suggested as a major new target for antibacterial and antifungal drug development [6,7]. The canonical two-component signal transduction system is shown in Figure 1B. A histidine kinase, often in response to receipt of a signal or stimulus, autophosphorylates on a conserved histidine residue. The phosphoryl group is then transferred to a conserved aspartate residue of a cognate response regulator. Phosphorylation of the response regulator occurs within the receiver domain and typically leads to a change in cellular physiology by activating an output domain. In many cases, phosphorylation enables the response regulator to bind DNA and function as a transcription factor. However, many other types of output domains are found that endow their response regulators with the ability to mediate protein–protein interactions or to perform enzymatic functions [8]. Two-component signaling pathways have been shown to respond to a wide range of stimuli, including sugars, peptides, antibiotics, and quorum-sensing signals. These signals trigger major physiological changes by changing programs of gene expression, altering swimming behavior, regulating proteolysis, or triggering differentiation [9,10].
Both histidine kinases and their targets, the response regulators, are easily identified in bacterial genomes solely by sequence homology. C. crescentus encodes 106 such proteins: 62 histidine kinases and 44 response regulators [11]. Some bacterial genomes encode as many as 250 of these signaling proteins, often amounting to more than 5% of all genes in a genome [12]. In Escherichia coli, the vast majority of two-component systems are encoded as operons of a histidine kinase and a response regulator that form an exclusive one-to-one phosphotransfer pair [13]. However, studies in C. crescentus and other bacteria reveal that two-component signaling pathways can often be highly branched, with many-to-one and one-to-many phosphotransfer relationships [5,14,15]. Such pathways are also often composed of kinases and regulators encoded in different operons scattered throughout a genome. In C. crescentus 41 histidine kinases and 19 response regulators, or 57% of all two-component genes, are orphans, not encoded in the same operon as another two-component gene. Identifying the connectivity of two-component signaling pathways is not possible by sequence analysis alone and is currently a major challenge. A recent report has attempted to map all such interactions in E. coli by systematically measuring phosphotransfer relationships between histidine kinases and response regulators [16].
Forward genetic screens in C. crescentus have identified 14 of the 106 two-component signaling genes as involved in cell cycle progression or differentiation (reviewed in [3,15]). The majority of these 14 are orphans and their connectivity remains poorly defined. Moreover, what role the other 92 two-component genes may play in regulating cell cycle progression and differentiation is largely unknown. Previous genetic screens may not have been saturated or may have had inherent biases, precluding identification of other important two-component regulators. To address these challenges, we undertook a systematic, comprehensive genetic and biochemical dissection of all 106 two-component signal transduction genes in the C. crescentus genome. Analysis of a complete set of deletion mutants identified 39 genes required for some aspect of growth or cell cycle progression, including nine essential genes. To identify phosphotransfer relationships, we developed a global in vitro biochemical approach that allows the identification of connections that are relevant in vivo. This technique takes advantage of data demonstrating that histidine kinases have an in vitro kinetic preference for their in vivo substrates. We demonstrate the utility of this integrated suite of systematic genetic and biochemical tools by identifying a previously unknown, but highly conserved, two-component pathway that is essential for growth of C. crescentus owing to a role in controlling cell envelope structure and integrity. The tools and approach presented can be applied to the study of two-component signaling proteins in other prokaryotes, including pathogens, and in any species having multiple two-component signaling systems, such as plants.
Results
Systematic Deletion of Two-Component Signaling Genes
We analyzed the C. crescentus genome and identified 106 genes that encode members of the two-component signal transduction family: 62 histidine kinases and 44 response regulators (for annotation procedures, see Materials and Methods). To begin comprehensive identification of two-component signaling pathways required for cell cycle progression, cell growth, or cell polarity in C. crescentus, we generated deletion strains for each of the histidine kinase and response regulator genes identified. Deletions were made using long-flanking homology constructs carried on suicide vectors and a two-step recombination process (Figure 2; Materials and Methods). Selection for tetracycline resistance ensures integration of the suicide vector, and growth on sucrose (sacB is lethal when sucrose is present in the medium) selects for plasmid excision and formation of a stable deletion strain (Figures 2A and Figures S1). The two-step deletion procedure allows rapid identification of essential genes. If a gene is essential, the second recombination event always fails, and stable deletions (tetracycline-resistant [tetR]/sucrose-resistant [sucroseR] colonies) cannot be recovered (Figures 2A and Figures S1). In such cases, all sucroseR/tetR colonies recovered are a result of sacB mutation, not loss of sacB.
Figure 2 Systematic Deletion of Two-Component Signal Transduction Genes
(A) Methodology used to generate chromosomal deletion strains. For each gene to be deleted, a suicide vector was constructed, with approximately 800-bp regions of homology upstream and downstream of the gene flanking a tetR cassette. See Materials and Methods and Figure S1 for details of plasmid construction. In a two-step process, deletion strains are isolated by selecting first for tetracycline resistance and then by sucrose counter-selection utilizing the sacB gene carried on the vector. Cells harboring the sacB gene die in the presence of sucrose. Hence, a deletion strain is identified as tetR/sucroseR. For nonessential genes, stable deletions are easily identified by screening 5–10 colonies after the two-step recombination. For essential genes, no tetR/sucroseR strains can be recovered (see text and Figure S1 for additional details).
(B and C) Swarm plate analysis of 97 nonessential two-component deletion strains. (B) Map of strain positions in the swarm plates. Wild-type CB15N is in positions A1 and J10 for comparison to mutant strains. (C) PYE swarm plate after 3 d of growth at 30 °C. Swarm sizes and densities were scored visually and digital images analyzed in Matlab (MathWorks, Natick, Massachusetts, United States). Strains exhibiting swarm plate phenotypes are listed in Table 2, except for ΔCC1221 in position E1, which is deleted for a kinase erroneously annotated as a histidine kinase.
Table 2 Nonessential Deletion Strains—Phenotypic Summary
We successfully generated stable deletion strains in rich medium (peptone yeast extract [PYE]) for 97 of the 106 C. crescentus two-component signaling genes. For these 97 genes, stable deletions were found after screening 5–10 colonies. For the remaining nine genes we tested at least 100 colonies after the final sucrose counter-selection (Figures 2A and Figures S1) and found that all still possessed the sacB gene, albeit inactivated. This suggests that each of these nine genes cannot be eliminated and hence each is essential for growth or viability (Table 1). This set includes all previously characterized essential two-component signal transduction genes in C. crescentus: ctrA, cckA, divK, and divL [17–21]. These results validate our method as a means to finding essential genes and strongly suggest that the five previously uncharacterized genes that could not be deleted (CC0530, CC1743, CC2931, CC2932, and CC3743) are also essential in C. crescentus.
Table 1 Essential Two-Component Signal Transduction Genes
CC0530 and CC3743 are both genes of unknown function. CC0530 encodes a predicted histidine kinase with two transmembrane domains and a periplasmic loop of about 130 amino acids. The protein encoded by CC3743 is a putative transcriptional regulator of the winged-helix OmpR subfamily (data not shown). CC2932 and CC2931 probably form an essential two-component pathway as orthologs of each are found in the same predicted operon, or adjacent operons, in a wide range of bacterial genomes. CC2931 encodes an ortholog of the response regulator PetR, which is essential in Rhodobacter capsulatus and required for oxidative respiration [22]. CC1743 is an ortholog of the gene ntrY, which may control growth in the presence of nitrate [23]. We thus suspected that CC1743 may be dispensable for growth in M2G minimal medium, where the sole nitrogen source is ammonium. Repeating the deletion procedure for CC1743 on minimal medium did in fact yield a stable deletion strain, so we classify CC1743 as a conditionally essential gene. We could not make any similar predictions for the other four new essential genes, and suggest that they are essential under most standard growth conditions.
Phenotypic Analysis of Nonessential Deletion Strains
We next examined the phenotypes of the 97 nonessential deletion strains using a swarm plate assay. Wild-type cells can swim through low-percentage agar, creating a large, circular colony, or swarm, via the combined effects of chemotaxis and growth. Defects in a number of processes, including cell motility, chemotaxis, growth, cell division, and cell cycle progression, can produce changes in swarm size or density. The swarm plate assay is thus a rapid, sensitive, and comprehensive method for initial phenotypic characterization. Each deletion mutant, as well as the wild-type CB15N, was innoculated into swarm plates made from rich (PYE) medium, and swarms were photographed after three days (Figure 2B and 2C). From digital images, swarm size and swarm density were quantified for each deletion strain relative to wild-type (Figure 2C). Of the 97 deletion strains, 30 exhibited a significantly altered swarm size or density (Table 2). Each of these genes was further characterized by measuring the log-phase generation time in rich medium and by examining cellular morphology for abnormalities in cell shape, cell length, motility, and stalk formation (Figure 3; Table 2).
Figure 3 Morphology of Selected Deletion Strains
Deletion strains were harvested at mid-log phase and imaged using differential interference contrast microscopy. Strains: (A) wild-type CB15N, (B) ΔCC0138, (C) ΔCC0744, (D) ΔCC0909, (E) ΔCC1063, (F) ΔCC2482, (G) ΔCC3315, and (H) ΔCC3471.
Strong candidates for cell cycle or cell growth regulatory genes are those marked by deletion strains that show a decrease in swarm size and a longer generation time. Five strains matching these criteria were found, including deletions of known cell cycle regulators (divJ and flbD) and three uncharacterized regulators (tacA, CC0138, and CC0744). Strains with smaller swarms but no change in generation time likely indicate genes required for motility or chemotaxis, and this group of genes includes the known chemotaxis (che) genes. Seven strains had larger and less dense swarms than wild-type, perhaps because of disruption of genes controlling the swarmer-to-stalked cell transition, leading to extended swimming.
In sum, the initial phenotypic characterization of our comprehensive library of two-component deletion strains has identified 39 genes (30 nonessential and nine essential)—or more than 35% of all two-component signaling genes—required for some aspect of growth, viability, morphogenesis, or cell cycle progression. This includes all 14 of the genes found by previous forward genetic screens for morphogenetic and cell cycle mutants (Tables 1 and 2), as well as 25 previously uncharacterized two-component signaling genes involved in regulating the C. crescentus life cycle. The uncharacterized genes are not simply those with subtle mutant phenotypes, as many have severe defects, including five that appear to be essential for growth or viability. Detailed characterization will be necessary to pinpoint the precise function of each of these uncharacterized genes.
Systematic Biochemical Analysis of Two-Component Signal Transduction
As a first step in further characterization of the two-component signaling genes involved in the cell cycle progression and development of C. crescentus, we sought to identify the response regulator targets of each histidine kinase. For orphan kinases and regulators, cognate pairs cannot easily be predicted based on sequence analysis alone. Of the 39 mutants showing phenotypes in the assays described above, 26 are orphans and their phosphotransfer pairings thus unknown. To systematically identify connectivity between two-component signaling proteins, we developed a global in vitro biochemical technique, which we term phosphotransfer profiling, to rapidly identify the targets of histidine kinases (Figure 4).
Figure 4 Phosphotransfer Profiling Method
(A) Phosphotransfer profile experiments involve three separate reactions: (1) autophosphorylation of the histidine kinase (HK) by radiolabeled ATP, (2) phosphotransfer to a response regulator (RR), and (3) dephosphorylation of the response regulator.
(B) Schematic of the phosphotransfer profiling technique. A single preparation of purified, autophosphorylated kinase (HK∼32P) is mixed with each response regulator from a given organism and analyzed for phosphotransfer by SDS-PAGE and autoradiography. The first lane shows a single band corresponding to the autophosphorylated histidine kinase and is used as a comparison for every other lane. Lanes 2–4 illustrate the three possible outcomes of a phosphotransfer reaction. In lane 2, phosphotransfer from HK to RR1 leads to the appearance of a band corresponding to RR1. In lane 3, phosphotransfer from HK to RR2 also occurs, but owing to high phosphatase activity (either autophosphatase or catalyzed by a bifunctional HK), the net result is production of inorganic phosphate (Pi) and the depletion of radiolabel from both the HK and RR2. In lane 4, no phosphotransfer occurs, and the lane is indistinguishable from lane 1.
(C–H) Phosphotransfer profiling was performed for three E. coli kinases (EnvZ, CheA, and CpxA) against all 32 purified E. coli response regulators, with phosphotransfer incubation times of either 1 h (C, E, and G) or 10 s (D, F, and H). For these three histidine kinases, a comparison of the short and long time point profiles indicates a kinetic preference for only their in vivo cognate regulators: OmpR (C and D), CheY and CheB (E and F), and CpxR (G and H). After being examined for phosphotransfer, all gels are stained with Coomassie to verify equal loading of histidine kinase and response regulator in each lane (data not shown). For each kinase profiled, we purified only its soluble, cytoplasmic domain, either as a thioredoxin-His6 or a His6-MBP fusion, using standard metal affinity chromatography (see Materials and Methods). When necessary, we made successive N-terminal truncations until we identified a construct that produced active kinase in vitro, always preserving the H-box and ATP binding domain (details on constructs used are in Table S3). All response regulators were purified as full-length fusions to a thioredoxin-His6 tag. Purity was assessed by Coomassie staining, with each purified kinase domain and response regulator, except for E. coli FimZ, yielding an intense band of the correct approximate molecular weight (see Figure S5; Table S3).
In a profiling experiment (Figure 4A and 4B), the purified cytoplasmic, soluble kinase domain of a histidine kinase is autophosphorylated with [γ-32P]ATP, and then split into separate reactions containing equimolar amounts of each purified, full-length response regulator (for details of protein purification, see Materials and Methods). Each phosphotransfer reaction is incubated for an identical period of time and then stopped by addition of sample buffer, separated by SDS-PAGE, and imaged on phosphor screens. A control of autophosphorylated kinase without any added response regulator is included for reference, and forms a single intense band. Efficient phosphotransfer to a response regulator can be manifested in two ways (Figure 4B). In the first case, a high-intensity band is seen at the appropriate molecular weight for phosphorylated response regulator. In the second case, efficient phosphotransfer can lead to depletion of radiolabel from the histidine kinase band. As some response regulators have high autophosphatase activity and some histidine kinases are bifunctional, also acting as specific phosphatases for their cognate response regulators, the net result of efficient phosphotransfer and phosphatase activities is the depletion of radiolabel from the autophosphorylated kinase (Figure 4A and 4B) [24]. Hence, to identify a phosphotransfer relationship, each reaction in a profile assay is inspected for (i) a band corresponding to the response regulator or (ii) a decrease in intensity of the kinase band relative to the kinase-only control. Importantly, because our profile method relies on the comparison, in parallel, of all potential phosphotransfer substrates for a given kinase, it is independent of the specific activity of the kinase being tested.
Histidine Kinases Exhibit a System-Wide In Vitro Kinetic Preference for Their Cognate Response Regulators
We chose to test and validate our in vitro profiling technique using purified kinases and response regulators from E. coli as many of its in vivo phosphotransfer pairings are known. First, we characterized phosphotransfer to response regulators by the histidine kinase EnvZ, which responds in vivo to changes in osmolarity by controlling the phosphorylation state of the response regulator OmpR [25,26]. The profile of EnvZ after a 1-h reaction time with each of the 32 purified E. coli response regulators demonstrates phosphotransfer to 11 different response regulators, including OmpR (Figure 4C). However, with a shorter, 10-s reaction time the only efficient phosphotransfer is to OmpR (Figure 4D), demonstrating a clear kinetic preference of EnvZ for its cognate substrate OmpR. We next tested the CheA histidine kinase, which phosphorylates CheY and CheB in vivo to control chemotaxis [27,28]. At 1 h, CheA shows phosphotransfer to seven response regulators, including CheY and CheB (Figure 4E), but at 10 s we detect only phosphorylation of CheY and CheB (Figure 4F). We then tested a third kinase, CpxA, which is known to signal through CpxR in vivo [29]. With the long reaction time, CpxA phosphorylates CpxR as well as several other response regulators (Figure 4G). The short reaction time again reveals a kinetic preference of the kinase CpxA for its in vivo, cognate substrate, CpxR (Figure 4H). We have observed similar kinetic preferences of two other E. coli kinases, PhoQ and PhoR, for their respective phosphotransfer substrates, PhoP and PhoB (data not shown). We conclude that E. coli histidine kinases have a strong kinetic preference for their in vivo cognate response regulators, with promiscuity only observed after extended incubation times. We have estimated the kinetic preference of kinases to be at least 103 in terms of relative k
cat/K
m ratios (Figure S2).
Next, we tested C. crescentus histidine kinases to determine if kinetic preference for substrates extends to the two-component systems in this organism. We started by profiling a two-component pair of unknown function: CC1181/CC1182. Because the kinase and regulator are encoded in the same operon they likely form an exclusive phosphotransfer pair in vivo. As with E. coli histidine kinases, we found that multiple response regulators were phosphorylated by CC1181 at the 1-h time point, including CC1182 (Figure 5A). However, a shorter phosphotransfer incubation time of 10 s revealed a clear kinetic preference of CC1181 for CC1182 (Figure 5B). We then tested five other C. crescentus histidine kinases, CC0289 (PhoR), CC0759, CC1740, CC2765, and CC3327. In each case, the histidine kinase exhibited a strong kinetic preference for its known substrate or the substrate encoded within its own operon, CC0294 (PhoB), CC0758, CC1741, CC2766, and CC3325, respectively (data not shown).
Figure 5 Phosphotransfer Profiling of C. crescentus Histidine Kinases
Profiles for four purified C. crescentus kinases versus 44 purified response regulators were obtained by the method described for E. coli in Figure 4.
(A) One-hour time point profile of the C. crescentus kinase CC1181.
(B) Ten-second time point profile. Only CC1182, encoded in the same operon as CC1181 and the likely in vivo target, is phosphorylated at the short time point. Kinetic preference of C. crescentus histidine kinases for their cognate substrates was similarly demonstrated for five other operon pairs (data not shown).
(C and D) Ten-second time point profiles of the orphan kinases DivJ and PleC, demonstrating phosphorylation of only their shared in vivo targets, PleD and DivK.
(E) Phosphotransfer profiling of the previously uncharacterized essential orphan kinase CC0530 (CenK) reveals a single preferred substrate, CC3743 (CenR).
Next, we used profiling with orphan C. crescentus histidine kinases for which the cognate response regulators could not be predicted by sequence analysis alone. First, we tested the orphan kinases DivJ and PleC, which were identified in our deletion analysis, and in previous genetic screens [30,31], to be key regulators of cell cycle progression and morphogenesis. Both of these kinases have been shown previously to phosphorylate the essential response regulator DivK and the response regulator PleD, which are in the same operon together, but without an adjacent kinase [20,21,32]. Short, 10-s reaction time profiles of DivJ and PleC demonstrate a kinetic preference for DivK and PleD and suggest that these are the exclusive targets of DivJ and PleC (Figure 5C and 5D).
We conclude that, as in E. coli, C. crescentus histidine kinases have an in vitro kinetic preference for their in vivo cognate substrate. Kinetic preference of a kinase for its cognate response regulator has been noted before on a limited scale [25,33–35], but our data extend this observation to a genome-wide level. Moreover, we suggest that this kinetic preference can be exploited to rapidly identify in vivo phosphotransfer relationships.
Identification of a New Essential Two-Component Pathway That Controls Cell Envelope Integrity
The systematic deletion analysis described above identified four histidine kinases that each appear to be essential for growth or viability: divL, cckA, CC2932, and CC0530 (see Table 1). divL and cckA have both been previously identified as essential regulators and are implicated in phosphorylation of the essential response regulator CtrA [18,19]. CC2932 is encoded in an operon with the essential response regulator CC2931, and these probably form a phosphotransfer pair. CC0530, however, is a previously uncharacterized, orphan kinase with no known or predicted substrate. Using phosphotransfer profiling, we demonstrated that CC0530 preferentially phosphorylates a single target, the orphan response regulator CC3743 (Figure 5E). As with CC0530, we had identified CC3743 as a previously uncharacterized orphan gene that is likely essential for growth or viability of C. crescentus (see Table 1). Together our genetic and biochemical observations strongly suggest that these two orphans comprise an essential two-component pathway in C. crescentus.
To test whether CC0530 and CC3743 are indeed essential, we generated strains in which the only copy of each gene is present on a low-copy plasmid under the control of the xylose-inducible, glucose-repressible promoter PxylX. For both genes, stable deletions were easily recovered when these complementing plasmids were present but not in the presence of an empty vector control (Table 3). This work produced strain ML521 (ΔCC0530 + PxylX-CC0530) and strain ML550 (ΔCC3743 + PxylX-CC3743). ML521 formed colonies only on plates supplemented with xylose, consistent with the CC0530 histidine kinase being essential for growth (data not shown). In contrast, ML550 formed colonies in the presence of xylose or glucose. We suspected that CC3743 may be a stable protein and hence difficult to deplete when expressed from a plasmid. We therefore made a destabilized version of CC3743 by adding a C-terminal ssrA tag, which targets proteins for degradation and decreases protein half-life inside the cell [36]. Using this destabilizing tag, we successfully created the strain ML591 (ΔCC3743 + PxylX-CC3743-ssrA), which forms colonies on PYE plates supplemented with xylose but not with glucose (data not shown). The ability of ML591 to grow on medium with xylose suggests that the ssrA tag does not interfere with the function of CC3743, but does allow the depletion of CC3743 during growth on glucose. The depletion strains ML521 and ML591 also grew only in minimal medium supplemented with xylose (M2X) and not with glucose (M2G), supporting the general essential nature of these two genes (data not shown).
Table 3 CC0530 (cenK) and CC3743 (cenR) Are Essential Genes
Next, we examined the phenotype of these strains in liquid medium after depleting each gene product. Cultures of each were grown in rich medium supplemented with xylose and then washed and resuspended at a low density in medium with xylose or glucose. We measured the growth rate and observed the cells by light microscopy (Figure 6A–6E). In the presence of xylose, growth of ML521 and ML591 was virtually indistinguishable from wild-type, suggesting that expression of either CC0530 or CC3743 under these conditions has no deleterious effect (Figure 6A). However, when shifted to glucose, the cultures of each depletion strain stopped growing and failed to accumulate significant optical density (Figure 6A). After ∼20 h of depletion by growth in glucose, we examined the morphological phenotype of each strain by light microscopy. Depletion of either gene product led to loss of motility, shorter stalks, and a dramatic, unusual membrane blebbing, resulting in bubble-like protrusions on the cell surface (Figure 6C and 6E). Cells were approximately wild-type in length and size, but had cell envelope blebs nearly covering the cell surface. We reasoned that the blebs were contiguous extrusions of the cell envelope that did not disrupt permeability as these cells did not lyse even after extended incubation in glucose-containing medium. Using high-resolution scanning electron microscopy, we examined cells from each depletion strain after extended growth in xylose and glucose. Consistent with the light microscopy results, we observed large, irregular protrusions across the surface of the cells grown in glucose and depleted of CC0530 or CC3743 (Figure 6F–6I). The growth and morphological phenotypes of the two depletion strains were nearly identical, further supporting the conclusion that CC0530 and CC3743 participate in the same signal transduction pathway. Based on our observations we have named CC0530 and CC3743 cenK (cell envelope kinase) and cenR (cell envelope regulator), respectively.
Figure 6 CC0530 (cenK) and CC3743 (cenR) Are Essential for Growth and Required for Cell Envelope Integrity
Growth curves for the ML521 (ΔCC0530 + PxylX-cenK) and ML591 (ΔCC3743 + pHXM-cenR-ssrA) depletion strains (A). Overnight cultures of each were grown in PYE plus xylose (PYE-X), washed with plain PYE, and diluted in PYE plus xylose or PYE plus glucose (PYE-G). After 12 h of growth in these conditions cells reached an optical density (OD600) level that could be measured (this time is plotted as “0 min”). Morphology was observed by light microscopy for the cenK depletion (ML521) after a total of 20 h in PYE plus xylose (B) or PYE plus glucose (C) and for the cenR depletion (ML591) after 20 h in PYE plus xylose (D) or PYE plus glucose (E). Scanning electron micrographs under identical conditions are shown for ML521 in PYE plus xylose (F) and PYE plus glucose (G) and for ML591 in PYE plus xylose (H) and PYE plus glucose (I). For (F–I), scale bar represents 1 μm. Depletion of either gene product led to an unusual, irregular blebbing of the cell surface. Cells were not motile, and had reduced stalk length.
To understand the functions of the cenK–cenR pathway in more detail we examined the effects of overexpressing components of this pathway (Figure 7). First we examined the phenotype of strain ML603, which expresses a full-length copy of cenR under control of the PxylX promoter on a low-copy plasmid (pMR20) in a wild-type background. In the presence of glucose, cells of this strain were virtually indistinguishable from wild-type cells (Figure 7A). However, in the presence of xylose, these cells showed significant cellular elongation, and many cells appeared to be losing their shape, exhibiting a bloated, enlarged morphology (Figure 7B). To increase expression further, we constructed a strain (ML675) with PxylX-cenR on pJS71, a higher-copy-number vector than pMR20. In the presence of glucose, strain ML675 also appeared similar to wild-type (Figure 7C), but growth in xylose revealed a dramatic morphological phenotype, ranging from bloated, enlarged cells to pervasive cell lysis (Figure 7D). Measurements of optical density after shift to xylose indicated a rapid growth arrest (Figure 7K). Interestingly, we noted that in many predivisional cells, the cell was enlarged asymmetrically, always with the stalked half of the cell losing its rod-like appearance (indicated by white arrows in Figure 7D). These data, together with the depletion analysis, suggest that cenR is involved in maintaining proper cell envelope structure, and further suggest that peptidoglycan or cell membrane synthesis may proceed in an asymmetric fashion in wild-type C. crescentus cells.
Figure 7 Constitutive Activation of the CenK–CenR Pathway Leads to Dramatic Changes in Cell Morphology, Cell Lysis, and Death
Images are shown for strains grown overnight in PYE plus glucose and then diluted back to early log phase and grown for 5 h in PYE plus glucose (A, C, E, G, and I) or xylose (B, D, F, H, and J). In all panels, white arrows indicate cells with asymmetric bloating and black arrows indicate lysed cells.
(A and B) ML603 (CB15N + pLXM-cenR + pJS71) expresses CenR alone from a low-copy vector.
(C and D) ML675 (CB15N + pHXM-cenR) expresses CenR alone from a high-copy vector.
(E and F) ML606 (CB15N + pLXM-cenR[D60E]) expresses, from a low-copy vector, a mutant of CenR that mimics constitutive phosphorylation.
(G and H) ML607 (CB15N + pMR20 + pHXM-cenK
cyto) expresses CenKcyto alone from a high-copy vector.
(I and J) ML604 (CB15N + pLXM-cenR + pHXM-cenK
cyto) expresses both CenR and CenKcyto, from low- and high-copy plasmids, respectively.
(K) Growth curve for all strains from (A–J) grown in PYE supplemented with xylose [64,65].
For many response regulators, mutating the conserved phosphorylation site from aspartate to glutamate mimics constitutive phosphorylation [37,38]. We introduced such a mutation, D60E, into cenR, on a low-copy plasmid. In the presence of glucose, the resulting cells looked similar to wild-type (Figure 7E), but when shifted to xylose, they became severely enlarged, lost their usual rod shape, and within 5 h began to lyse and die (Figure 7F and 7K). Thus, the phenotype of overexpressing CenR(D60E) on a low-copy plasmid matched that of overexpressing wild-type CenR on a high-copy plasmid (compare Figure 7D and 7F). We conclude that the D60E mutation leads to phosphorylation-independent activity of CenR. We also attempted to generate strains expressing CenR(D60E) from the high-copy plasmid pJS71, but no colonies were recovered, even on glucose plates, suggesting that the D60E allele may be so active that even basal expression in glucose is lethal.
Unlike with CenR, overexpression of the full-length CenK (data not shown) or its cytoplasmic kinase domain had no effect on cell growth or cell morphology (Figure 7G and 7H). This may be because the amount of CenR is limiting in the cell, so that additional CenK expression may not alter the fraction of phosphorylated CenR. Alternatively, the cell may be robust to changes in kinase concentration, as suggested for the kinase EnvZ [39]. Regardless, we predicted that if CenK is the in vivo cognate kinase for CenR, then simultaneously overexpressing both CenK and CenR should phenocopy overexpression of CenR(D60E). As expected, the effect of co-overexpressing CenKcyto and CenR (Figure 7J) was significantly more severe than that of expressing either protein alone (compare to Figure 7B and 7H), and phenocopied the overexpression of CenR(D60E) (Figure 7F). As a control to ensure that the effect was due to kinase activity of CenK, we mutated the active-site histidine to alanine (H273A) and showed that the growth rate of cells co-overexpressing CenK(H273A) and CenR was nearly indistinguishable from that of cells overexpressing CenR alone (data not shown). These results support the conclusion that CenK acts in vivo to phosphorylate, and hence activate, CenR, as suggested by the in vitro phosphotransfer profiling.
The CenK–CenR pathway appears to be widely conserved throughout the alpha subdivision of proteobacteria. Multiple sequence alignments indicate better than 60% identity (70% similarity) for CenR and better than 35% identity (50% similarity) for CenK (Figures S3 and S4). The similarity extends throughout the full length of each protein, including the putative periplasmic ligand-binding domain of CenK. We suggest that the CenK–CenR pathway may be essential and function similarly in a range of other bacteria.
Discussion
Systematic Deletion of Two-Component Signal Transduction Genes
By deleting each of the 106 two-component signal transduction genes encoded in the C. crescentus genome, we have identified 39 mutant strains with cell cycle or developmental phenotypes (see Tables 1 and 2). Previous forward genetic screens had identified 14 two-component signaling genes involved in cell cycle progression and morphogenesis in C. crescentus, including four essential for viability of the organism. However, forward genetic screens are typically designed to select for a particular phenotype or may not be screened to saturation. The comprehensive, unbiased nature of the reverse genetic approach taken here expands both the number and role of two-component signaling proteins in regulating the C. crescentus cell cycle. The newly identified mutants include many with severe phenotypes as well as four previously uncharacterized genes that appear to be essential for growth or viability in both rich and minimal media. The library of deletion strains created here will also serve as a resource for future explorations of two-component regulation in C. crescentus. The deletion strains can be individually characterized in more depth, in different conditions, or even in different strain backgrounds. In addition, the inclusion of unique molecular bar codes in each strain (see Materials and Methods) opens the possibility of parallel fitness studies similar to those used for the Saccharomyces cerevisiae whole-genome deletion collection [40,41].
Systematic Biochemical Analysis of Two-Component Phosphorylation
Similarity of mutant phenotypes can help to identify two-component genes acting in the same pathway, but ultimately, a biochemical demonstration of phosphorylation is required to define signal transduction pathways. Such a combination of genetics and biochemistry has successfully defined individual two-component signaling pathways in a number of organisms [10], but this report presents a global, integrated genetic and biochemical study of a bacterium's complete set of two-component signal transduction systems.
Histidine kinases have been widely thought to function promiscuously in vitro, precluding correspondence with in vivo targets. However, a few studies have suggested that histidine kinases may have a kinetic preference in vitro for their in vivo cognate substrates. For example, in Bacillus subtilis, the kinase KinA can phosphorylate both Spo0A and Spo0F in vitro, but has a more than 50,000-fold preference, as measured by relative k
cat/K
m ratios, for Spo0F, its in vivo cognate substrate [35]. A similar magnitude of kinetic preference was shown for the kinase VanS phosphorylating its cognate regulator VanR relative to the noncognate substrate PhoB [34]. The phosphotransfer profiling data presented here extend these observations to a system-wide level and suggest that the apparent promiscuity of histidine kinases in vitro is attributable to excessive incubation times or a high concentration of reaction components, each of which acts to cross the kinetic barrier that enables a kinase to selectively phosphotransfer to its cognate substrate. A recent comprehensive study of two-component signal transduction in E. coli examined phosphotransfer in vitro from each histidine kinase to each response regulator at a 30-s time point [16]. As with our data, all known cognate pairs showed significant phosphotransfer, but the study reported a small number of interactions between noncognate pairs [16]. However, the in vivo relevance of these interactions is not yet known, and because that study did not examine phosphotransfer at multiple time points, the strength of noncognate interactions relative to those of cognate pairs is also not yet clear.
Our profiling method examines, simultaneously and in parallel, the ability of a purified histidine kinase to phosphorylate each of the response regulators encoded in that organism's genome. It would be impractical to determine k
cat/K
m for each kinase–regulator combination, but kinetic preference can still easily be seen by conducting comprehensive profiles at multiple time points. Importantly, using a number of previously well-characterized E. coli histidine kinases, we demonstrated a direct correspondence between this kinetic preference and in-vivo-relevant response regulator substrates (see Figure 4). We were then able to use this kinetic preference to identify in vivo targets of uncharacterized histidine kinases such as the C. crescentus orphan CenK (see Figure 5E). Note, however, that phosphotransfer profiling is not used in isolation to identify phosphotransfer pairs, but is integrated with genetic data and in vivo experiments, as demonstrated here for CenK–CenR.
The phosphotransfer profiling technique is robust to a number of experimental variables. First, it is independent of the specific activity of the purified histidine kinase, because the method relies on a relative comparison of phosphotransfer kinetics from a single preparation of kinase to each possible substrate. Second, because the kinetic preference of kinases appears to be on the order of 103 or even 104, the method is not significantly affected by differences in response regulator concentration, even differences as great as 10-fold. Also, some histidine kinases are bifunctional, acting as both a kinase and a phosphatase for their cognate response regulators. In most cases, control of the relative ratio is not understood in vivo, making it difficult to predict the ratio of kinase to phosphatase activity of a particular purified construct in vitro. Any construct having net kinase activity can be profiled by our method to identify the probable in vivo substrates, but determining whether the histidine kinase acts predominantly as a kinase or a phosphatase in vivo depends on integration with genetic and other in vivo observations. For example, our profiles of DivJ and PleC, as well as previous studies of these kinases, suggest that both target the regulators DivK and PleD [20,21,32]. In vivo, though, DivJ is thought to function primarily as a kinase for DivK and PleD, whereas the bifunctional kinase PleC appears to act as a phosphatase [42,43].
Identifying Novel Signal Transduction Systems
We demonstrated the integration of our genetic and biochemical methods to identify a novel, essential pathway from the histidine kinase CenK to the response regulator CenR, which appears to control critical aspects of cell envelope integrity. CenK is a predicted transmembrane protein with a periplasmic domain of ∼130 amino acids, although no periplasmic stimulus could be predicted based on sequence. CenR is a predicted DNA-binding protein of the OmpR subfamily, so defining the CenR regulon may help to unveil its role in controlling the cell envelope. Depletion of either gene product led to a severe membrane blebbing phenotype, which, to the best of our knowledge, has not been seen before in C. crescentus. A number of other C. crescentus genes are involved in maintaining cell wall integrity and cell shape, including mreB, rodA, and cicA, but the relationship, if any, of these genes to cenK and cenR is not yet clear [44–46].
CenK–CenR is, to our knowledge, the first essential two-component pathway discovered in Gram-negative bacteria controlling cell envelope processes. In some Gram-positive bacteria, an essential two-component pathway, YycG–YycF, also plays a role in cell envelope biogenesis [47–49], but does not appear to be orthologous to the CenK–CenR system. However, the CenK–CenR pathway does appear to be highly conserved throughout the alpha subdivision of proteobacteria, including a number of important plant, animal, and human pathogens. Two-component systems have been highlighted as a possible new antibiotic target given their absence in humans and other animals [6,7,50,51]. Furthermore, as the physical construction of the cell wall has long been a major target of antibiotics, the CenK–CenR regulatory pathway may be a particularly suitable target for novel antibiotic development.
Signaling Pathway Specificity and Insulation
All organisms use a relatively small number of signaling modalities. For bacteria such as C. crescentus two-component signaling systems are widely employed, whereas eukaryotes have large families of other signaling systems, such as MAP kinase cascades, TGF-β pathways, and receptor tyrosine kinases. By definition, cross-talk between pathways must be minimal, otherwise an organism would be unable to trigger specific responses to specific stimuli. However, the mechanisms and strategies employed by cells to insulate highly related pathways are poorly understood and have been a recent focus of attention in many organisms [52–55].
We propose that the system-wide kinetic preference of histidine kinases for their cognate response regulators is a fundamental mechanism by which bacterial cells maintain the insulation of two-component signaling pathways. The large kinetic preference of kinases for their cognate substrates suggests that cross-talk observed in vitro likely arises from excesses in reaction time or reaction components and does not occur in vivo. Importantly, we distinguish deleterious cross-talk from cross-regulation in which a single kinase has multiple bona fide targets or multiple kinases regulate the same response regulator. There are several well-studied examples of cross-regulation, such as the E. coli kinase CheA, which phosphorylates both CheY and CheB as part of its role in regulating chemotaxis [27], and some of the noncognate interactions found in a systematic study of E. coli two-component signaling may represent additional cases of cross-regulation [16]. In C. crescentus, cross-regulation occurs between the orphan kinases DivJ and PleC, and the two response regulators DivK and PleD. Our profile data for CheA, DivJ, and PleC demonstrated that kinases involved in cross-regulation have approximately equal kinetic preference for their multiple response regulator targets (see Figure 5C and 5D).
There are, of course, many additional means by which cells ensure signaling specificity. For example, subcellular localization of interacting components, scaffolding, and mutual inhibition can all act to ensure specificity [53]. However, our in vitro results point to biochemical selectivity as a fundamental mechanism, on which other layers of regulation and insulation may be built. Recent results with the cyclin-dependent kinases suggest that biochemical selectivity may also play a fundamental role in this process in S. cerevisiae [54]. It remains a major challenge to understand in complete detail how organisms robustly and accurately ensure signal fidelity within a cell [55].
Concluding Remarks
The techniques and approach described here can be directly extended to any organism containing two-component signal transduction systems, and are particularly useful for species with large sets of these molecules. This includes most bacteria, which typically encode at least 20 or 30 two-component genes and sometimes more than 100. Many plant species, including the model system Arabidopsis thaliana and the agriculturally and economically important rice plant Oryza sativa, also contain large sets of two-component signaling genes.
Finally, all cells, even relatively simple bacteria, are capable of complex information-processing tasks, such as converting continuous signals to discrete outputs, signal amplification, coincidence detection, and cellular-level memory. The successful implementation of these tasks is not carried out by individual proteins, but rather by multiple proteins, arranged into complex, highly connected circuits. For example, MAP kinase pathways are capable of converting continuous signals to an all-or-none output owing to a precise connectivity, a three-tiered MAPK cascade, and positive feedback [56]. Mapping the structure of signaling pathways and networks, as initiated here for C. crescentus, will thus be critical to our understanding of how cells process information and make decisions in order to regulate their behavior.
Materials and Methods
Bacterial strains, plasmids, and growth conditions
E. coli strains were routinely grown in Luria Broth (BD Biosciences, Franklin Lakes, New Jersey, United States) at 37 °C, supplemented with carbenicillin (100 μg ml−1 or 50 μg ml−1), chloramphenicol (30 μg ml−1 or 20 μg ml−1), kanamycin (50 μg ml−1 or 30 μg ml−1), oxytetracycline (12 μg ml−1), or spectinomycin (50 μg ml−1) as needed for solid and liquid media. C. crescentus strains were grown in PYE (complex medium) or M2G (minimal medium) at 30 °C [57]. PYE medium was supplemented with 3% sucrose, oxytetracycline (2 μg ml−1 or 1 μg ml−1), kanamycin (25 μg ml−1 or 5 μg ml−1), or spectinomycin (100 μg ml−1 or 25 μg ml−1), as required. PYE swarm plates contained 0.3% bacto agar. Site-directed mutagenesis of cenK and cenR was carried out using the primers CenKH273Afw, CenKH273Arev, CenRD60Efw, and CenRD60Erev, using the QuikChange protocol (Stratagene, La Jolla, California, United States). pKOC3 was constructed by PCR amplification of the tetR cassette from pMR20 using the primers tet-fw and tet-rev, digestion with EcoRI, and ligation into the EcoRI site of pBluescript. Strains, plasmids, and primers used in this study are listed in Tables 4 and Figures S1–Figures S3.
Table 4 Strains and Plasmids
Deletion of C. crescentus two-component genes
Response regulators and histidine kinases were identified by BLAST analysis of the C. crescentus genome sequence using known two-component protein sequences as input. For response regulators, sequences with BLAST E-values less than 0.01 were inspected for presence of the conserved residues D12, D13, D57, T87, and K109, where numbering is for E. coli CheY [10]. In sum, 44 response regulators were identified; these include two which may not be phosphorylated owing to mutation of one of the five highly conserved residues: CC3100 and CC0612. For histidine kinases, sequences with BLAST E-values less than 0.01 were inspected for presence of the conserved H-, N-, D/F-, and G-boxes [10]. Two histidine kinases, CC0433 and CC0594, are CheA-like and have a P1 domain instead of the usual H-box. Nine histidine kinases are members of the newly identified HWE group [58] and lack the F-box (CC0629, CC0836, CC1683, CC2554, CC2909, CC3048, CC3058, CC3170, and CC3560).
Deletion strains were generated by a long-flanking homology procedure and two-step recombination (see Figure S1) [59,60]. Complete lists of primers used are in Table S2. Regions of homology flanking each gene to be deleted were amplified in 50-μl reactions by PCR using the following conditions: 40 ng CB15N genomic DNA, 50 μM each dNTP, 100 nM each primer (P1 + P2a or P3a + P4), 1X Pfu Turbo buffer, 1.25 U Pfu Turbo polymerase (Stratagene), 2% DMSO, and 60 mM Betaine. For each reaction, 35 cycles of the following sequence were run: 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 5 min. Reactions included a pre-incubation at 94 °C for 5 min, and concluded with a 10-min extension at 72 °C. Products were then re-amplified using identical conditions, but with primers P1 and P2b or P3b and P4. This produced final regions of flanking homology that were gel-purified (Qiagen, Valencia, California, United States) and used to amplify a tetR cassette by PCR: 50 μM each dNTP, 100 nM P1 primer, 100 nM P4 primer, the products of the flanking homology PCRs, 1 mM MgCl2, 1X Taq buffer, 2% DMSO, 60 mM Betaine, 2.5 U Taq (Invitrogen, Carlsbad, California, United States), 0.5 U Pfx (Invitrogen), and 200 ng of the KpnI-SacI fragment of pKOC3 containing the tetR cassette. Cycling comprised pre-incubation at 94 °C for 2 min; followed by ten cycles of the sequence 94 °C for 30 s, 53 °C for 30 s, and 72 °C for 8 min; followed by 20 cycles of the sequence 94 °C for 30 s, 53 °C for 30 s, and 72 °C for 8 min plus 20 additional seconds per cycle; followed by 72 °C for 10 min. Final PCR amplicons were gel-purified, blunted using the End-IT kit (Epicentre, Madison, Wisconsin, United States) and ligated into pNPTS138. Ligations were transformed into DH5α and positive colonies selected by blue/white screening. Plasmids from white colonies were verified by restriction digest with BamHI and HindIII or by sequencing. We term these plasmids “knockout plasmids” and name each according to the nomenclature pKO-CCXXXX, where CCXXXX is the unique GenBank identifier for the gene to be deleted. Knockout plasmids were transformed into CB15N by electroporation, and first integrants selected by plating on PYE containing oxytetracycline. Colonies were innoculated into liquid PYE medium with oxytetracycline and grown for 12–16 h. Five microliters of each culture was then plated on PYE plates containing oxytetracycline and sucrose. Colonies were screened for tetracycline resistance and kanamycin sensitivity to identify deletion strains. Proper construction of the gene deletion was also verified by two PCRs: one used tet-conf and a primer specific to the chromosomal region of the deleted gene (Pconf) and the other used sacBfw and sacBrev to verify loss of the sacB gene.
The first six and last 12 codons of each gene deleted were left intact to protect against disruption of possible regulatory signals for adjacent genes. For genes at the beginning of an operon, the presence of a tetR cassette may lead to polar effects, but the majority of mutants found to have phenotypes (see Figure 2; Tables 1 and 2) were encoded as single genes and hence would not exhibit polarity. For simplicity we refer to all mutant strains by the gene disrupted, whether in a predicted operon or not. For mutants of interest that may suffer from polar effects, the tetR cassette can be removed, as done for CC0530 and CC3743 (see below). The cassette is flanked by two direct repeats (FRT sites) such that expression of the FLP recombinase catalyzes removal of the tetR cassette, leaving behind an in-frame deletion construct [61]. Each deletion strain was also engineered to incorporate two unique 20mer bar codes, each of which is not found elsewhere in the C. crescentus genome. The bar code sequences were adapted from the S. cerevisiae deletion project [40,41], enabling similar high-throughput phenotypic characterization of deletion strains using high-density microarrays with probes complementary to the bar codes (M. T. L., unpublished data).
Generation of pENTR clones for response regulators and histidine kinases
Strains for expression and purification of His6-tagged proteins were generated using the Gateway high-throughput recombinational cloning system (Invitrogen). For each response regulator, the entire gene was amplified by PCR, using reverse phase cartridge purified primers (Sigma-Genosys, St. Louis, Missouri, United States). PCR reactions contained 60 mM Betaine, 2% DMSO, 1X Pfu buffer, 50 μM each dNTP, 75 ng CB15N genomic DNA, 10 pmol each primer, and 1.25 U Pfu Turbo (Stratagene). Reactions were incubated at 95 °C for 5 min, followed by 35 cycles of 95 °C for 1 min, 55 °C or 58 °C for 1 min, and 72 °C for 5 min, and finished by a 10-min extension at 72 °C. PCR amplicons were cloned into the pENTR/D-TOPO vector according to the manufacturer's protocol, and transformed into TOP10 competent cells (Invitrogen). Kanamycin-resistant (kanR) colonies were screened by colony PCR using M13F and M13R to verify the correct insert size. Positive clones were sequence-verified using M13F and M13R. In total, 76 Gateway adapted response regulator pENTR clones were generated for this study (32 for E. coli and 44 for C. crescentus). Each clone generated is named pENTR-CCXXXX or pENTR-bXXXX, where CCXXXX and bXXXX are the unique GenBank identifiers for the C. crescentus and E. coli genes, respectively. Cloning of histidine kinases was done identically to that of the response regulators, but the forward primer was designed to eliminate any transmembrane domains predicted by the SMART database (http://smart.embl-heidelberg.de). For complete primer lists, see Table S3.
Destination vectors and recombinational cloning
Expression vectors were constructed and adapted for recombinational cloning using the Gateway vector conversion system (Invitrogen). Two plasmids (pTRX-HIS-DEST and pHIS-MBP-DEST) were derived from the IPTG-inducible pET32a and pET15b vectors (Novagen. Madison, Wisconsin, United States). To construct pTRX-HIS-DEST, a MscI-XhoI fragment of pBADM-20 (EMBL protein purification and expression facility) was used to replace the same region of pET32a. This resulting clone was digested with NcoI and blunted with T4 DNA polymerase, and the RfA Gateway cassette was cloned into this site. To construct pHIS-MBP-DEST, an XbaI-BamHI fragment of pETM-41 (EMBL protein purification and expression facility) was first cloned into pET15b. The resulting clone was digested with NcoI and blunted with T4, and the RfA cassette was cloned into this filled-in site. The vectors were designed to generate an N-terminal fusion to thioredoxin-His6, or His6-maltose binding protein, followed by the TEV protease cleavage site. Two other destination vectors were constructed for inducible expression in C. crescentus based on the low-copy (pMR20) and high-copy (pJS71) plasmids. These plasmids utilize the promoter region of the xylX gene [62] and add an N-terminal M2 epitope tag (DYKDDDDK) to the gene of interest immediately after the start methionine. To construct pHXM-DEST, the xylX promoter region was amplified with the primers XYLM2fw and XYLM2rev and cloned into pJS71 as a SacI-SalI fragment. This clone was then digested with SalI and blunted with T4, and the RfB cassette (Invitrogen) was cloned into this site. pLXM-DEST was derived from pHXM-DEST by removing a SacI-KpnI fragment, and blunt cloning into the SacI site of pMR20. For both of these vectors, PxylX is in the opposite direction of the PlacZ promoter. Using Gateway LR clonase reactions to mediate site-specific recombination, pENTR response regulator and histidine kinase clones were recombined with these destination vectors to create expression clones for either protein purification or in vivo C. crescentus studies. Each 10-μl LR reaction contained: 50 ng of destination vector, ∼50 ng of pENTR plasmid DNA, 1X LR buffer, 3 U topoisomerase I, and 1 μl of LR clonase enzyme mix (Invitrogen). Reactions were incubated overnight at room temperature, transformed into chemically competent DH5α cells, and plated on LB with antibiotics as necessary. Colonies were tested for resistance to ampicillin (pTRX-HIS-DEST and pHIS-MBP-DEST), spectinomycin (pHXM-DEST), or tetracycline (pLXM-DEST), tested for sensitivity to kanamycin to ensure no carryover of pENTR DNA, and PCR verified with T7F and T7R for E. coli expression vectors or M13F and M13R primers for C. crescentus vectors. The resulting expression plasmids were called pTRX-HIS-CCXXXX, for the C. crescentus response regulators and pTRX-HIS-bXXXX, for the E. coli response regulators. Expression plasmids for the E. coli and C. crescentus histidine kinases have a parallel nomenclature.
Protein expression and purification
Expression plasmid DNA was transformed into E. coli BL21-Tuner cells. Single colonies were grown in 500 ml of LB to OD600 ∼0.6 and fusion proteins induced by addition of 300 μM IPTG. Cells were grown at 37 °C prior to induction and then shifted to 30 °C for 4 h before harvesting by centrifugation at 10,800 g for 5 min. Cells were pelleted and stored at −80 °C until needed. Native purifications of His6-tagged proteins were performed using affinity chromatography with Ni-NTA agarose beads (Qiagen). All steps of the purification (except for elution) were performed in batch using 50-ml conical tubes. The following buffers were used for purification: lysis buffer (20 mM Tris-HCl [pH 7.9], 0.5 M NaCl, 10% glycerol, 20 mM imidazole, 0.1% Triton X-100, 1 mM PMSF, 1 mg/ml lysozyme, 125 units benzonase nuclease [Novagen]), wash buffer (20 mM HEPES-KOH [pH 8.0], 0.5 M NaCl, 10% glycerol, 20 mM imidazole, 0.1% Triton X-100, 1 mM PMSF), elution buffer (20 mM HEPES-KOH [pH 8.0], 0.5 M NaCl, 10% glycerol, 250 mM imidazole), and storage buffer (10 mM HEPES-KOH [pH 8.0], 50 mM KCl, 10% glycerol, 0.1 mM EDTA, 1 mM DTT). Each cell pellet was resuspended in 10 ml of lysis buffer, incubated at room temperature for 20 min, sonicated, and then centrifuged for 60 min at 30,000 g to generate a cleared lysate. His6-tagged proteins were bound to 1 ml of Ni-NTA agarose slurry, washed twice with 50 ml of wash buffer, and then loaded onto an Econo-column (Bio-Rad, Hercules, California, United States) for elution. Purified protein was eluted using 2.5 ml of elution buffer and loaded directly onto a PD-10 column (Amersham Biosciences, Piscataway, New Jersey, United States) that had been pre-equilibrated with storage buffer. If necessary, samples were filtered with a 0.2-μm HT Tuffryn filter (Pall Gelman Sciences, East Hills, New York, United States), and then concentrated to approximately 1–10 mg/ml using Centricon YM-10 or YM-30 columns (Millipore, Billerica, Massachusetts, United States). All samples were filtered through an Ultrafree-MC (0.22 μm) spin filter (Millipore) and then aliquoted for storage at −80 °C. Protein concentrations were measured using Coomassie Plus Protein Assay Reagent and a BSA standard (Pierce Biotechnology, Rockford, Illinois, United States). An equal amount (500 ng) of each protein sample was analyzed by 12% SDS-PAGE to verify molecular weight and purity. Prior to phosphotransfer profiling, all response regulator concentrations were normalized against a 500-ng BSA standard using a ChemiImager 5500 and densitometry (Alpha Innotech, San Leandro, California, United States) (see Figure S5).
Phosphotransfer profiling
Each purified kinase was autophosphorylated in storage buffer supplemented with 2 mM DTT, 5 mM MgCl2, 500 μM ATP, and 5 μCi [γ32P]ATP (∼6,000 Ci/mmol, Amersham Biosciences). Reactions were allowed to proceed until equilibrium at 30 °C (15 min to 2 h depending on the kinase). Purified response regulators were diluted to a final concentration of 5 μM in storage buffer plus 5 mM MgCl2. Phosphotransfer reactions contained 5 μl of phosphorylated kinase and 5 μl of response regulator (2.5 μM final concentration of each) and were incubated at 30 °C. Reactions were stopped with 3.5 μl of 4X sample buffer (500 mM Tris [pH 6.8], 8% SDS, 40% glycerol, 400 mM β-mercaptoethanol) and stored on ice until loaded. The entire sample was loaded, without heating, on 10% Tris-HCl ready gels (Bio-Rad) and electrophoresed at room temperature for 50 min at 150 V. The dye front and unincorporated ATP was removed with a razor blade and the wet gel (still on the back glass plate) placed in a Ziploc bag and exposed to a phoshor screen for 1−3 h at room temperature. The screen was scanned with a Storm 860 imaging system (Amersham Biosciences) at 50 μm resolution. E. coli profiles consisted of three protein gels, which were scanned separately and the images stitched together for analysis. C. crescentus profiles consisted of four protein gels, and were analyzed in the same fashion.
Estimation of kinetic preference
To estimate kinetic preference, we purified radiolabeled kinase by repeated washing with a Nanosep-30K column (Pall, East Hills, New York, United States). Autophosphorylation and phosphotransfer reactions were as described for phosphotransfer profiling, except that response regulators were diluted in storage buffer plus 5 mM MgCl2 plus 0.5 mg/ml bovine serum albumin. The final concentrations of kinase and regulator in the reaction were 2.5 μM and 0.25 μM, respectively. Kinetics of phosphotransfer were determined by quantifying bands using ImageQuant software (Amersham Biosciences). The fraction of phosphorylated response regulator was calculated by normalizing to the intensity of the band corresponding to kinase alone. These normalized values were plotted versus reaction time and used to estimate initial reaction velocities for cognate versus noncognate substrates.
Depletion, overexpression, and coexpression strains
A xylose-inducible low-copy plasmid was generated by amplifying the xylX promoter region with XYLSACfw and XYLNCOrev and cloning into pMR20, to generate pMR20-PxylX. This plasmid contains a unique NcoI site engineered at the start codon of the xylX gene. We then amplified, by PCR, full-length versions of CC0530 (cenK) and CC3743 (cenR) flanked by NcoI and HindIII sites using the primers CenKNCOfw, CenKH3rev, CenRNCOfw, and CenRH3rev. The full-length cenK and cenR PCR products were cloned into pMR20-PxylX, to generate pMR20-PxylX-cenK and pMR20-PxylX-cenR.
Next, we produced in-frame derivatives of pKO-CC0530 and pKO-CC3743 (see above). Each of these plasmids was cotransformed into E. coli with pCP20, which contains an arabinose-inducible FLP recombinase gene [61]. Expression of the FLP recombinase, according to the methods of Datsenko and Wanner [61], led to recombination between the direct repeat FRT sites flanking the tetR cassette. The resulting plasmids, pΔcenK-IF and pΔcenR-IF were sequenced to verify formation of an in-frame, markerless deletion construct. CB15N was then electroporated with pΔcenK-IF and pΔcenR-IF to generate kanR, sucrose-sensitive integrants. These first integrants were made electrocompetent and transformed with the complementing plasmids pMR20-PxylX-cenK or pMR20-PxylX-cenR described above, or a vector control pMR20. Single colonies from each transformation were then grown overnight in PYE containing oxytetracycline and 0.03% xylose (for cenK) or 0.0003% xylose (for cenR). After overnight growth, 1 μl was plated for counter-selection on PYE containing 3% sucrose and oxytetracycline. Using markerless deletion constructs, there were three possible outcomes for colonies from counter-selection: (i) regeneration of the wild-type allele, (ii) generation of an in-frame deletion, or (iii) sacB inactivation without plasmid excision. Ninety-six sucroseR colonies were picked, tested for kanamycin sensitivity, and analyzed by PCR using CenKconf_fw plus CenKconf_rev or CenRconf_fw plus CenRconf_rev. Colonies that were kanamycin sensitive and yielded a single PCR band of the expected size indicated colonies with an in-frame deletion. These procedures produced the depletion strains ML521 and ML550. ML521 grew only in the presence of xylose. As strain ML550 formed single colonies on plates supplemented with either glucose or xylose, we constructed a destablized version of CenR by adding a C-terminal ssrA tag (AANDNFAEEFAVAA) using the primers CenRfw and CenRssrArev. This construct was cloned into pENTR/D-TOPO and recombined into pHXM-DEST by the LR clonase reaction, to generate pHXM-cenR-ssrA. This plasmid was used to construct strain ML591 (ΔcenR + pHXM-cenR-ssrA), as described above for ML550. ML591 grew only in the presence of xylose.
To study the depletion strains ML521 and ML591, overnight cultures grown in xylose (0.03% for ML521 and 0.3% for ML591) were washed twice with PYE and diluted back 1:20,000 in PYE plus 0.1% glucose for CenK or 1:1,000 for CenR. As a control, each strain was also diluted 1:20,000 in PYE plus xylose. After 12 h of depletion, cells had grown to a sufficient density to be measured, and OD600 was monitored for an additional 8 h before samples were fixed for light and electron microscopy.
For overexpression and coexpression studies, CenKcyto, CenR, CenR(D60E), and CenKcytoH273A expression vectors were generated by Gateway cloning (see Tables 4 and S1). These vectors were electroporated into CB15N and selected with oxytetracyline, spectinomycin, or both as necessary. Overnight cultures were washed in PYE and diluted 1:50 in PYE supplemented with glucose or xylose. Samples were monitored by OD600 and fixed for light microscopy. Overexpression of CenR on pHXM-cenR was performed with 0.3% xylose, whereas all other experiments were done using 0.03% xylose.
Microscopy
C. crescentus cells were grown to mid-log phase, fixed by addition of 0.5% paraformaldehyde in PBS, washed, and concentrated with PYE medium. Samples were deposited on microscope slides coated with 0.1% poly-L-lysine. Differential interference contrast images were obtained with a Zeiss (Oberkochen, Germany) Axioskop2 equipped with a 100× Plan-NEOFLUAR (NA 1.3) objective and an AxioCam monochrome CCD camera controlled by Axiovision 4.1 software. For field emission scanning electron microscopy (FESEM), cells were pelleted at 10,000 g and resuspended in fix solution (0.1 M sodium cacodylate buffer [pH 7.4], 2% glutaraldehyde, 0.5% paraformaldehyde, 7.5% sucrose). After concentrating, cells were deposited on a poly-L-lysine-coated glass coverslip, then post-fixed with 1% osmium tetroxide in 0.1 M sodium cacodylate buffer and 7.5% sucrose. After washing, samples were dehydrated with an ethanol series (10 min each of 50%, 70%, 90%, 95%, and 100%), critical-point dried by the CO2 method (AutoSamdri 850, Tousimis, Rockville, Maryland, United States), and sputter coated with an approximately 5-nm layer of gold/palladium (Desk II, Denton Vacuum, Moorestown, New Jersey, United States). Cells were imaged with a LEO 982 FESEM (Zeiss) using a SE-INLENS detector operated at 2.0 keV.
Supporting Information
Figure S1 Diagram of Two-Step Deletion Procedure
Deletion constructs were generated with a splice-overlap extension protocol using six different primers (see Table S2). Each gene is disrupted by a tetR cassette and is flanked by approximately 800 bp of upstream (LFH) and downstream (RFH) flanking homologous DNA for efficient recombination. A suicide vector for each gene to be deleted contains a kanR gene and a sucrose counter-selectable sacB marker. A two-step recombination procedure results in the generation of a chromosomal deletion strain (tetR, kanS, sucroseR). This method also allows the identification of putative essential genes (tetR, kanR, sucroseR).
(6.5 MB TIF).
Click here for additional data file.
Figure S2 Estimation of Kinetic Preference
(A and B) Time courses for phosphorylation of OmpR and CpxR by EnvZ. In our phosphotransfer profiling (Figure 4C), OmpR and CpxR were both phosphorylated at the 60-min time point, but only OmpR was phosphorylated at the 10-s time point.
(C) Plot of normalized PhosphorImager counts for OmpR and CpxR phosphorylation based on a quantification of the gels shown in (A) and (B). Initial velocities (v0) were determined by measuring the slope (counts/second) for OmpR between 0 and 5 s, and for CpxR between 0 and 4,000 s.
(D and E) Time courses for phosphorylation of CC1182 and CC2931 by CC1181. In our profiling, both CC1182 and CC2931 were phosphorylated by CC1181 at 60 min, but only CC1181 was phosphorylated at 10 s.
(F) Plot of normalized PhosphorImager counts for CC1182 and CC2931 phosphorylation based on a quantification of the gels shown in (D) and (E). Initial velocities were determined for CC1182 between 0 and 10 s, and for CC2931 between 0 and 4,000 s. See Materials and Methods for experimental details. To estimate kinetic preference, in the Michaelis–Menton formalism, at substrate concentrations much less than K
m: v0 ≍ [E][S] (k
cat/K
m). In these time courses, we tested response regulators at a concentration of 0.25 μM, which is approximately 10-fold lower than the typical K
m of a kinase–regulator pair [63]. As the enzyme and substrate concentrations used in these time courses were identical, the ratio of k
cat/K
m for a cognate substrate relative to a noncognate substrate is estimated simply by: v0,congate/v0,non-cognate. From a quantification of our time course data, this ratio is approximately 1,330 for OmpR relative to CpxR and approximately 1,760 for CC1182 relative to CC2931. An in-depth kinetic characterization of individual kinases would be necessary to precisely determine k
cat and K
m, but the order of magnitude is 103. Moreover, in each case, this 103-fold preference is for a cognate response regulator relative to the next best substrate, suggesting that other regulators are separated by an even greater kinetic gap.
(4.0 MB TIF).
Click here for additional data file.
Figure S3 Multiple Sequence Alignment of CenR Orthologs
Putative CenR orthologs were identified by reciprocal best BLAST analysis. CenR proteins are highly conserved in the alpha subdivision of proteobacteria (C. crescentus CB15, Agrobacterium tumefaciens C58, Sinorhizobium meliloti 1021, Mesorhizobium loti MAFF303099, Brucella melitensis 16M, Rhodopseudomonas palustris CGA009, Bradyrhizobium japonicum USDA 110, Rhodobacter sphaeroides 2.4.1, Silicibacter pomeroyi DSS-3).
(5.4 MB TIF).
Click here for additional data file.
Figure S4 Multiple Sequence Alignment of CenK Orthologs
Putative CenK orthologs were identified by reciprocal best BLAST analysis. CenK proteins are highly conserved in the alpha-subdivision of proteobacteria (C. crescentus CB15, Agrobacterium tumefaciens C58, Sinorhizobium meliloti 1021, Mesorhizobium loti MAFF303099, Brucella melitensis 16M, Rhodopseudomonas palustris CGA009, Bradyrhizobium japonicum USDA 110, Rhodobacter sphaeroides 2.4.1, Silicibacter pomeroyi DSS-3).
(2.3 MB TIF).
Click here for additional data file.
Figure S5 Purified C. crescentus and E. coli Response Regulators
(A) Thirty-two E. coli response regulators were purified as thioredoxin-His6 fusion proteins.
(B) Fourty-four C. crescentus response regulators were purified as thioredoxin-His6 fusion proteins. Approximately 500 ng of purified protein was analyzed by 12% SDS-PAGE. The predicted molecular weights can be found in Table S3. Only one response regulator, E. coli FimZ, was not purified in a soluble form (no band of the correct weight is found in this lane). A molecular weight ladder is labeled in kilodaltons.
(3.2 MB TIF).
Click here for additional data file.
Table S1 Primer Names and Sequences Used for Plasmids Constructed in This Study
(15 KB XLS).
Click here for additional data file.
Table S2 Primers for Deletion of C. crescentus Two-Component Signal Transduction Genes
For each gene to be deleted, six primers were required (P1, P2a, P2b, P3a, P3b, and P4) plus one gene-specific confirmation primer (Pconf) (see Figure S1). The resulting deletion constructs are called “pKO-CCXXXX” where CCXXXX is the unique GenBank identifier number.
(66 KB XLS).
Click here for additional data file.
Table S3 Primers for pENTR Clones of Histidine Kinases and Response Regulators
List of primers used to clone 44 C. crescentus response regulators, and 32 E. coli response regulators. Each resulting pENTR clone is called pENTR-CCXXXX or pENTR-bXXXX for C. crescentus and E. coli genes, respectively. Three E. coli histidine kinases and four C. crescentus histidine kinases were also cloned, and the primers used are listed.
(33 KB XLS).
Click here for additional data file.
Accession Numbers
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank/) accession numbers for the CenR orthologs discussed in this paper are Agrobacterium tumefaciens C58 (Atu2763), Bradyrhizobium japonicum USDA 110 (bll0620), Brucella melitensis 16M (BMEI0066), C. crescentus CB15 (CC3743), Mesorhizobium loti MAFF303099 (MBNC03001136), Rhodobacter sphaeroides 2.4.1 (Rsph03000729), Rhodopseudomonas palustris CGA009 (RPA0283), Silicibacter pomeroyi DSS-3 (STM1w01002705), and Sinorhizobium meliloti 1021 (SMc03820). GenBank accession numbers for the CenK orthologs are Agrobacterium tumefaciens C58 (Atu0388), Bradyrhizobium japonicum USDA 110 (bll8095), Brucella melitensis 16M (BMEI1648), C. crescentus CB15 (CC0530), Mesorhizobium loti MAFF303099 (MBNC03004238), Rhodobacter sphaeroides 2.4.1 (Rsph03002719), Rhodopseudomonas palustris CGA009 (RPA0635), Silicibacter pomeroyi DSS-3 (STM1w01001404), and Sinorhizobium meliloti 1021 (SMc01716).
We thank Kathleen Ryan, Harley McAdams, Kurt Thorn, Laura Garwin, and Andrew Murray for helpful discussions and comments on the manuscript. We also thank Richard Schalek at the Center for Nanoscale Systems at Harvard University for assistance in scanning electron microscopy.
We gratefully acknowledge support from the Office of Science (BER), U.S. Department of Energy, grant numbers DE-FG03-01ER63219 and DE-FG02-04ER63922. Support was also provided in part by a National Institutes of Health grant to MTL at the Bauer Center for Genomics Research.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. JMS, MSP, BSP, EGB, and MTL conceived, designed, performed, and analyzed the experiments. JMS and MTL wrote the paper.
¤ Current address: Department of Molecular and Cell Biology, University of California, Berkeley, California, United States of America
Citation: Skerker JM, Prasol MS, Perchuk BS, Biondi EG, Laub MT (2005) Two-component signal transduction pathways regulating growth and cell cycle progression in a bacterium: A system-level analysis. PLoS Biol 3(10): e334.
Abbreviations
kanRkanamycin resistant/resistance
PYEpeptone yeast extract
sucroseRsucrose-resistant/resistance
tetRtetracycline resistant/resistance
==== Refs
References
Ryan KR Shapiro L Temporal and spatial regulation in prokaryotic cell cycle progression and development Annu Rev Biochem 2003 72 367 394 12651741
McAdams HH Shapiro L A bacterial cell-cycle regulatory network operating in time and space Science 2003 301 1874 1877 14512618
Skerker JM Laub MT Cell-cycle progression and the generation of asymmetry in Caulobacter crescentus
Nat Rev Microbiol 2004 2 325 337 15031731
Iba H Fukuda A Okada Y Rate of major protein synthesis during the cell cycle of Caulobacter crescentus
J Bacteriol 1978 135 647 655 681285
Stock AM Robinson VL Goudreau PN Two-component signal transduction Annu Rev Biochem 2000 69 183 215 10966457
Barrett JF Hoch JA Two-component signal transduction as a target for microbial anti-infective therapy Antimicrob Agents Chemother 1998 42 1529 1536 9660978
Stephenson K Hoch JA Two-component and phosphorelay signal-transduction systems as therapeutic targets Curr Opin Pharmacol 2002 2 507 512 12324251
Galperin MY Nikolskaya AN Koonin EV Novel domains of the prokaryotic two-component signal transduction systems FEMS Microbiol Lett 2001 203 11 21 11557134
Inouye M Dutta R Histidine kinases in signal transduction 2003 San Diego Academic Press 520
Hoch JA Silhavy TJ Two-component signal transduction 1995 Washington (D.C.) ASM Press 488
Nierman WC Feldblyum TV Laub MT Paulsen IT Nelson KE Complete genome sequence of Caulobacter crescentus
Proc Natl Acad Sci U S A 2001 98 4136 4141 11259647
Peterson JD Umayam LA Dickinson T Hickey EK White O The Comprehensive Microbial Resource Nucleic Acids Res 2001 29 123 125 11125067
Blattner FR Plunkett G Bloch CA Perna NT Burland V The complete genome sequence of Escherichia coli K-12 Science 1997 277 1453 1474 9278503
Bijlsma JJ Groisman EA Making informed decisions: Regulatory interactions between two-component systems Trends Microbiol 2003 11 359 366 12915093
Ohta N Grebe TW Newton A Brun YV Shimkets LJ Signal transduction and cell cycle checkpoints in developmental regulation of Caulobacter
Prokaryotic development 2000 Washington (D.C.) ASM Press 341 359
Yamamoto K Hirao K Oshima T Aiba H Utsumi R Functional characterization in vitro of all two-component signal transduction systems from Escherichia coli
J Biol Chem 2005 280 1448 1456 15522865
Quon KC Marczynski GT Shapiro L Cell cycle control by an essential bacterial two-component signal transduction protein Cell 1996 84 83 93 8548829
Jacobs C Domian IJ Maddock JR Shapiro L Cell cycle-dependent polar localization of an essential bacterial histidine kinase that controls DNA replication and cell division Cell 1999 97 111 120 10199407
Wu J Ohta N Zhao JL Newton A A novel bacterial tyrosine kinase essential for cell division and differentiation Proc Natl Acad Sci U S A 1999 96 13068 13073 10557274
Hecht GB Lane T Ohta N Sommer JM Newton A An essential single domain response regulator required for normal cell division and differentiation in Caulobacter crescentus
Embo J 1995 14 3915 3924 7664732
Wu J Ohta N Newton A An essential, multicomponent signal transduction pathway required for cell cycle regulation in Caulobacter
Proc Natl Acad Sci U S A 1998 95 1443 1448 9465034
Tokito MK Daldal F
petR located upstream of the fbcFBC operon encoding the cytochrome bc1 complex, is homologous to bacterial response regulators and necessary for photosynthetic and respiratory growth of Rhodobacter capsulatus
Mol Microbiol 1992 6 1645 1654 1323023
Pawlowski K Klosse U de Bruijn FJ Characterization of a novel Azorhizobium caulinodans ORS571 two-component regulatory system, NtrY/NtrX, involved in nitrogen fixation and metabolism Mol Gen Genet 1991 231 124 138 1661370
Dutta R Yoshida T Inouye M The critical role of the conserved Thr247 residue in the functioning of the osmosensor EnvZ, a histidine Kinase/Phosphatase, in Escherichia coli
J Biol Chem 2000 275 38645 38653 10973966
Igo MM Ninfa AJ Stock JB Silhavy TJ Phosphorylation and dephosphorylation of a bacterial transcriptional activator by a transmembrane receptor Genes Dev 1989 3 1725 1734 2558046
Igo MM Ninfa AJ Silhavy TJ A bacterial environmental sensor that functions as a protein kinase and stimulates transcriptional activation Genes Dev 1989 3 598 605 2663643
Hess JF Oosawa K Kaplan N Simon MI Phosphorylation of three proteins in the signaling pathway of bacterial chemotaxis Cell 1988 53 79 87 3280143
Ninfa EG Stock A Mowbray S Stock J Reconstitution of the bacterial chemotaxis signal transduction system from purified components J Biol Chem 1991 266 9764 9770 1851755
Raivio TL Silhavy TJ Transduction of envelope stress in Escherichia coli by the Cpx two-component system J Bacteriol 1997 179 7724 7733 9401031
Sommer JM Newton A Turning off flagellum rotation requires the pleiotropic gene pleD: pleA, pleC and pleD define two morphogenic pathways in Caulobacter crescentus
J Bacteriol 1989 171 392 401 2536661
Sommer JM Newton A Pseudoreversion analysis indicates a direct role of cell division genes in polar morphogenesis and differentiation in Caulobacter crescentus
Genetics 1991 129 623 630 1752411
Paul R Weiser S Amiot NC Chan C Schirmer T Cell cycle-dependent dynamic localization of a bacterial response regulator with a novel di-guanylate cyclase output domain Genes Dev 2004 18 715 727 15075296
Burbulys D Trach KA Hoch JA Initiation of sporulation in B. subtilis is controlled by a multicomponent phosphorelay Cell 1991 64 545 552 1846779
Fisher SL Kim SK Wanner BL Walsh CT Kinetic comparison of the specificity of the vancomycin resistance VanS for two response regulators, VanR and PhoB Biochemistry 1996 35 4732 4740 8664263
Grimshaw CE Huang S Hanstein CG Strauch MA Burbulys D Synergistic kinetic interactions between components of the phosphorelay controlling sporulation in Bacillus subtilis
Biochemistry 1998 37 1365 1375 9477965
Keiler KC Shapiro L TmRNA is required for correct timing of DNA replication in Caulobacter crescentus
J Bacteriol 2003 185 573 580 12511504
Smith JG Latiolais JA Guanga GP Pennington JD Silversmith RE A search for amino acid substitutions that universally activate response regulators Mol Microbiol 2004 51 887 901 14731287
Klose KE Weiss DS Kustu S Glutamate at the site of phosphorylation of nitrogen-regulatory protein NTRC mimics aspartyl-phosphate and activates the protein J Mol Biol 1993 232 67 78 8331671
Batchelor E Goulian M Robustness and the cycle of phosphorylation and dephosphorylation in a two-component regulatory system Proc Natl Acad Sci U S A 2003 100 691 696 12522261
Giaever G Chu AM Ni L Connelly C Riles L Functional profiling of the Saccharomyces cerevisiae genome Nature 2002 418 387 391 12140549
Shoemaker DD Lashkari DA Morris D Mittmann M Davis RW Quantitative phenotypic analysis of yeast deletion mutants using a highly parallel molecular bar-coding strategy Nat Genet 1996 14 450 456 8944025
Aldridge P Paul R Goymer P Rainey P Jenal U Role of the GGDEF regulator PleD in polar development of Caulobacter crescentus
Mol Microbiol 2003 47 1695 1708 12622822
Wheeler RT Shapiro L Differential localization of two histidine kinases controlling bacterial cell differentiation Mol Cell 1999 4 683 694 10619016
Fuchs T Wiget P Osteras M Jenal U Precise amounts of a novel member of a phosphotransferase superfamily are essential for growth and normal morphology in Caulobacter crescentus
Mol Microbiol 2001 39 679 692 11169108
Figge RM Divakaruni AV Gober JW MreB, the cell shape-determining bacterial actin homologue, co-ordinates cell wall morphogenesis in Caulobacter crescentus
Mol Microbiol 2004 51 1321 1332 14982627
Wagner JK Galvani CD Brun YV
Caulobacter crescentus requires RodA and MreB for stalk synthesis and prevention of ectopic pole formation J Bacteriol 2005 187 544 553 15629926
Ng WL Kazmierczak KM Winkler ME Defective cell wall synthesis in Streptococcus pneumoniae R6 depleted for the essential PcsB putative murein hydrolase or the VicR (YycF) response regulator Mol Microbiol 2004 53 1161 1175 15306019
Martin PK Li T Sun D Biek DP Schmid MB Role in cell permeability of an essential two-component system in Staphylococcus aureus
J Bacteriol 1999 181 3666 3673 10368139
Fabret C Hoch JA A two-component signal transduction system essential for growth of Bacillus subtilis Implications for anti-infective therapy J Bacteriol 1998 180 6375 6383 9829949
Furuta E Yamamoto K Tatebe D Watabe K Kitayama T Targeting protein homodimerization: a novel drug discovery system FEBS Lett 2005 579 2065 2070 15811319
Watanabe T Hashimoto Y Yamamoto K Hirao K Ishihama A Isolation and characterization of inhibitors of the essential histidine kinase, YycG in Bacillus subtilis and Staphylococcus aureus
J Antibiot (Tokyo) 2003 56 1045 1052 15015732
Zarrinpar A Park SH Lim WA Optimization of specificity in a cellular protein interaction network by negative selection Nature 2003 426 676 680 14668868
Schwartz MA Madhani HD Principles of MAP kinase signaling specificity in Saccharomyces cerevisiae
Annu Rev Genet 2004 38 725 748 15568991
Loog M Morgan DO Cyclin specificity in the phosphorylation of cyclin-dependent kinase substrates Nature 2005 434 104 108 15744308
Elion EA Qi M Chen W Signal transduction. Signaling specificity in yeast Science 2005 307 687 688 15692041
Ferrell JE Machleder EM The biochemical basis of an all-or-none cell fate switch in Xenopus oocytes Science 1998 280 895 898 9572732
Ely B Genetics of Caulobacter crescentus
Methods Enzymol 1991 204 372 384 1658564
Karniol B Vierstra RD The HWE histidine kinases, a new family of bacterial two-component sensor kinases with potentially diverse roles in environmental signaling J Bacteriol 2004 186 445 453 14702314
Wach A PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in S. cerevisiae
Yeast 1996 12 259 265 8904338
Schweizer HP Hoang TT An improved system for gene replacement and xylE fusion analysis in Pseudomonas aeruginosa
Gene 1995 158 15 22 7789804
Datsenko KA Wanner BL One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products Proc Natl Acad Sci U S A 2000 97 6640 6645 10829079
Meisenzahl AC Shapiro L Jenal U Isolation and characterization of a xylose-dependent promoter from Caulobacter crescentus
J Bacteriol 1997 179 592 600 9006009
Yoshida T Cai S Inouye M Interaction of EnvZ, a sensory histidine kinase, with phosphorylated OmpR, the cognate response regulator Mol Microbiol 2002 46 1283 1294 12453215
Laub MT McAdams HH Feldblyum T Fraser CM Shapiro L Global analysis of the genetic network controlling a bacterial cell cycle Science 2000 290 2144 2148 11118148
Evinger M Agabian N Envelope-associated nucleoid from Caulobacter crescentus stalked and swarmer cells J Bacteriol 1977 132 294 301 334726
|
16176121
|
PMC1233412
|
CC BY
|
2021-01-05 08:21:29
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e334
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030334
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1616784610.1371/journal.pbio.0030338Research ArticleBiotechnologyCell BiologyMolecular Biology/Structural BiologyNeurosciencePhysiologyDermatologyBiochemistryHomo (Human)Mus (Mouse)The Period Length of Fibroblast Circadian Gene Expression Varies Widely among Human Individuals Circadian Gene Expression in HumansBrown Steven A [email protected]
1
Fleury-Olela Fabienne
1
Nagoshi Emi
2
Hauser Conrad
3
Juge Cristiana
4
Meier Christophe A
4
Chicheportiche Rachel
3
Dayer Jean-Michel
3
Albrecht Urs
5
Schibler Ueli [email protected]
1
1 Department of Molecular Biology, University of Geneva, Geneva, Switzerland,2 Department of Biology—Howard Hughes Medical Institute, Brandeis University, Waltham, Massachusetts, United States of America,3 Division of Immunology and Allergy, University Hospital Geneva, Geneva, Switzerland,4 Endocrine Unit, Laboratory of Molecular Endocrinology, University Hospital Geneva, Geneva, Switzerland,5 Department of Medicine, Section of Biochemistry, Fribourg, SwitzerlandMignot Emmanuel Academic EditorStanford UniversityUnited States of America10 2005 27 9 2005 27 9 2005 3 10 e33814 9 2004 3 8 2005 Copyright: © 2005 Brown 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.
Fibroblast Culture Cells Give Scientists the Time of Day
Mammalian circadian behavior is governed by a central clock in the suprachiasmatic nucleus of the brain hypothalamus, and its intrinsic period length is believed to affect the phase of daily activities. Measurement of this period length, normally accomplished by prolonged subject observation, is difficult and costly in humans. Because a circadian clock similar to that of the suprachiasmatic nucleus is present in most cell types, we were able to engineer a lentiviral circadian reporter that permits characterization of circadian rhythms in single skin biopsies. Using it, we have determined the period lengths of 19 human individuals. The average value from all subjects, 24.5 h, closely matches average values for human circadian physiology obtained in studies in which circadian period was assessed in the absence of the confounding effects of light input and sleep–wake cycle feedback. Nevertheless, the distribution of period lengths measured from biopsies from different individuals was wider than those reported for circadian physiology. A similar trend was observed when comparing wheel-running behavior with fibroblast period length in mouse strains containing circadian gene disruptions. In mice, inter-individual differences in fibroblast period length correlated with the period of running-wheel activity; in humans, fibroblasts from different individuals showed widely variant circadian periods. Given its robustness, the presented procedure should permit quantitative trait mapping of human period length.
A novel method permits characterization of circadian rhythms in humans and mice using single skin biopsies.
==== Body
Introduction
Circadian rhythms of physiology and behavior in mammals are dependent upon a central clock that resides in the suprachiasmatic nucleus (SCN) of the brain hypothalamus. This clock is synchronized to the outside world via light input from the retina, and it in turn entrains similar slave oscillators present in most cells of the body [1]. In constant darkness, the circadian clock will direct sleep–wake cycles and many other physiological processes according to its intrinsic period length, which may be longer or shorter than 24 h. Because the clock is reset by light each day, its intrinsic period length influences the relative phase of circadian physiology and activity patterns. Thus, in human beings there is a correlation between circadian period length and the entrained phase of physiological rhythms and sleep-wake timing [2,3]. Extremely early and late activity patterns are thought to be associated with advanced and delayed sleep phase syndromes, respectively. Both advanced and delayed sleep phase syndromes can have genetic causes, and polymorphisms in three circadian clock genes, CK1ɛ, PER2, and PER3, have been linked to or associated with cases of familial advanced or delayed sleep phase syndromes [4–6]. Polymorphisms in the latter gene have also been associated more generally with diurnal preference [7].
The characterization of human clocks and their genetic defects is rendered challenging by the difficulty and expense of measuring human circadian period, since prolonged subject observation under laboratory conditions is required. In mice, the period length of circadian behavior is determined by analysis of wheel-running behavior in constant darkness. Recently, however, it has been possible to complement mouse behavioral analyses by measuring the period length of circadian gene expression in vitro from transgenic animals in which the luciferase gene has been fused to a circadian promoter [8,9]. For these animals, circadian rhythms were analyzed in explants from different tissues simply by real-time measurement of light output. Using the same technology, high-amplitude circadian gene expression can also be measured in cultured mouse NIH 3T3 fibroblasts whose oscillators are synchronized through a short treatment with serum or dexamethasone, a glucocorticoid receptor agonist [10]. Moreover, single-cell recordings of cultured mouse and rat fibroblasts have demonstrated that the circadian oscillators of these cells are self-sustained and cell-autonomous [10,11], similar to those operative in SCN neurons [12,13]. The circadian rhythms of electrical firing frequencies of dissociated individual SCN neurons display considerable intercellular differences in period length (τ). However, the mean τ-values determined for neuron populations harvested from wild-type and tau mutant hamsters closely correlate with the ones measured for the locomotor activity of these animals [13]. Hence, the genetic makeup of the clockwork circuitry appears to influence cellular and behavioral oscillations in a similar fashion.
A method of measuring human circadian rhythms from tissue biopsies would greatly complement behavioral studies of circadian rhythms and the disorders affecting them, since genetic differences appear to manifest themselves in both central and peripheral oscillators [14,15]. In this paper, we employed a lentivirally delivered circadian reporter shielded by enhancer-blocking activities to achieve this result. The distribution of period lengths that we measured from 19 human subjects demonstrates that inter-individual genetic differences in circadian clock function can be measured in skin biopsies. In mice, clock function measured in this way correlated with the period of wheel-running behavior. In both organisms, the wide range of fibroblast circadian period lengths obtained suggests interesting differences between physiology controlled by the SCN and circadian gene expression directed by peripheral oscillators in vitro.
Results
To measure circadian rhythms from a single biopsy, a circadian reporter must be introduced into the cells of a cultured tissue sample. However, human primary cultures do not easily permit transient transfection. Moreover, transiently transfected cells typically contain very high numbers of introduced reporter genes, an imbalance that can alter normal circadian rhythms by titrating circadian regulatory proteins [16]. Therefore, we turned to lentiviral delivery as a method of introducing a stably integrated construct in low copy number to primary fibroblast cultures [17]. We developed a lentivirus that contains a luciferase gene whose expression is governed by the promoter and 3′ untranslated regions of the mouse circadian gene Bmal1. To test this virus, it was used to infect immortalized 3T3 fibroblasts, and then circadian rhythms in these cells were synchronized by dexamethasone treatment [18]. BMAL1-luciferase expression was subsequently measured in the cell population by real-time recording of light output [9]. Cells infected with this virus gave high levels of expression, but very low circadian amplitude, so it was useless for circadian period measurements (data not shown).
Because lentiviruses integrate preferentially into the coding regions of active genes [19], we reasoned that the circadian behavior of the Bmal1 promoter was hampered by interference from loci at which the virus integrated or, less likely, by viral sequences themselves. To shield the reporter gene from such influences, we introduced multimerized FII insulator sequences from the chick β-globin gene [20] upstream and downstream of the Bmal1 reporter. These sequences have been previously shown to possess enhancer-blocking activity in vivo. When 3T3 fibroblasts were infected as above with these modified viruses, robust circadian oscillations were observed. As a parallel strategy, we introduced an “enhancer trap decoy,” consisting of the strong promoter of the human Elongation Factor 1α (EF1α) gene immediately followed by an SV40 transcription terminator, upstream of the Bmal1 promoter. The resultant “decoy” construct also yielded excellent circadian oscillations of luciferase activity (Figure 1A and 1B). Because signal magnitude was consistently greater with it than with the insulated construct, it was used for the experiments described in this paper. To insure that the period lengths of the oscillations observed using this virus were not affected by the titer of virus used or by the degree of infection, 3T3 cells were infected with various amounts of virus, and circadian rhythms measured as above. The period length of the oscillations was identical in all cases, although signal amplitude varied over a wide range (Figure 1C).
Figure 1 Circadian Bioluminescence Can Be Recorded in Fibroblasts Infected with a Lentiviral Luciferase Expression Vector
(A) Circadian reporter constructs used in these studies. Each contains the mouse Bmal1 promoter, the firefly luciferase coding region, and the Bmal1 3′UTR, flanked by the long terminal repeats (LTRs) of a lentiviral packaging vector. In (i), a dimerized chick β-globin FII element is inserted between each LTR and adjacent Bmal1 sequences. In (ii), a DNA segment composed of the EF1α promoter and a SV40 terminator is inserted between the upstream LTR and Bmal1 promoter, and the gfp coding region between the Bmal1 UTR and the downstream viral LTR.
(B) 3T3 cells were infected with the lentiviral vectors shown above, and equivalent infection levels were verified by real-time PCR to detect integrated viruses. Four days after infection, cells were shocked with dexamethasone to synchronize circadian rhythms, and luciferase output was measured by real-time luminometry.
(C) 3T3 cells were infected with different concentrations of the lentiviral reporter vector ii (see [A]), and circadian rhythms were measured as in (B). 10× represents unconcentrated filtered viral supernatant, 1× represents a 10× dilution of this, and 100× was a 10× concentration by ultracentrifugation. The number of viral infection units/plate were approximately 10,000 (1×), 30,000 (3×), 100,000 (10×), 300,000 (30×), and 1,000,000 (100×).
Next, for two different individuals this virus was used to infect 50,000 activated human monocytes purified from a single blood donation, or 50,000 human fibroblasts amplified from a single 2-mm skin punch biopsy (see Materials and Methods for details). After 4 d, cellular rhythms were synchronized with dexamethasone, and luciferase output was measured. In both cell populations, circadian oscillations were observed; but with skin fibroblasts, we obtained much higher signals and greater amplitudes of circadian oscillation (Figure 2A and 2B); hence, more precise period lengths could be estimated. On rare occasions, it has also been possible to cultivate hair root keratinocytes that cling to the end of a plucked human hair. These keratinocytes can also be infected with lentivirus, and give period lengths identical to those from fibroblasts of the same subject (Figure 2C). However, because most plucked hairs do not contain keratinocytes without performing scalp biopsies, we decided to continue our analysis of human circadian rhythms using fibroblasts isolated from normal skin biopsies.
Figure 2 Circadian Bioluminescene Recordings from Primary Human Blood Cells, Fibroblasts, and Keratinocytes
(A) For two different individuals, 50,000 non-immortalized adult primary human skin fibroblasts were infected with lentiviral reporter vectors. Four days after infection, cells were shocked with dexamethasone, and circadian rhythms were measured by real-time luminometry.
(B) For two different individuals, 50,000 human blood monocytes were treated as in (A).
(C) A skin biopsy was taken from one individual, and 50,000 fibroblasts were treated as in (A) and (B). In parallel, hairs were plucked until a hair was withdrawn that contained hair root keratinocytes clinging to the proximal end. These were cultivated and amplified to 50,000 cells, then infected and measured identically to fibroblasts.
To measure human circadian rhythms, two to five 2-mm diameter skin biopsies were taken from the abdomen or buttocks of 12 healthy normal individuals. Four additional human fibroblast populations were obtained from male foreskin, and three from other sources (see Materials and Methods for details.). From each of these 19 samples, 50,000 adult skin fibroblasts were infected with reporter virus, and circadian rhythms were measured as described previously. Two measurements on two infected populations from each biopsy were taken. Four sample curves are shown in Figure 3A, and the data are summarized in Figure 3B. An average period length of fibroblast circadian gene expression of 24.5 h was obtained, with a standard deviation of 45 min.
Figure 3 Circadian Bioluminescence Cycles in Fibroblasts from Different Human Individuals
Biopsies were obtained from buttocks, foreskin, or abdomen of 19 individuals (see Materials and Methods for details). Fibroblasts were isolated from each biopsy, infected with lentiviral circadian reporter vectors as in Figure 2, and analyzed by real-time luminometry. Individuals are designated with the letters A–S.
(A) Representative BMAL1-luciferase oscillations measured from biopsies of four different individuals. Individuals N, L, A, and P are shown.
(B) Summary of the period lengths of BMAL-luciferase oscillations from all 19 individuals. Each value shows the average plus or minus the standard deviation from two different trials of two different infections of fibroblasts from two to five biopsies per subject. The probability by Student's t-test that the most different individuals (A and S) have the same period length is ≤0.00001; the probability that the second most different (B and R) are equal is ≤0.004.
(C) For four subjects from whom two to five biopsies were taken, the average plus or minus the standard deviation of period length from two infections and four measurements of each skin biopsy is shown. The probability that the individuals that differ the most (C and R) do not differ in period length is ≤0.000002.
The period length of different cultures could in principle vary from biopsy to biopsy, or it could vary from individual to individual and remain constant among different biopsies of the same individual. Obviously, only in the latter case would the results be diagnostically useful. Hence, it was important to compare the range of data from different biopsies of the same individual with the range of data from different individuals. In multiple cases, inter-individual differences were significantly greater than the differences observed between cultures; four such examples and their statistical analyses are described in Figure 3C. Overall, the standard deviation among different trials using the same sample was 18 min; among samples derived from different infections of the same sample, the standard deviation was 25 min; and among the average values of different biopsies from the same person, the standard deviation was 6 mins. We conclude that this method can detect small differences in fibroblast circadian period length. What is particularly fascinating, however, is that the standard deviation among different individuals in our trial was 48 min. Thus, significant genetic differences in fibroblast clock function exist even in very small population samples.
To ensure that circadian genetic differences are indeed reflected in the rhythms of fibroblast gene expression that we measure, we applied this reporter system to measure circadian rhythms from tail biopsies of mice containing several known circadian mutations that shorten, lengthen, or abolish the period of circadian wheel-running behavior. Table 1 lists the mouse strains that we used and the published properties of their circadian clocks. We obtained adult dermal fibroblasts from tail biopsies of each of these nearly isogenic mice, and analyzed their circadian rhythms exactly as done in humans. Parallel to this analysis, circadian wheel running was measured for the same individuals (Figure 4). Mice with a period of wheel-running behavior shorter than wild-type (Per1brdm/brdm) yielded fibroblasts whose period of circadian Bmal1 expression was also shorter. Similarly, fibroblasts from mice with a period of wheel running that was longer than wild-type (Per1brdm/brdm;Cry2−/−, and Cry2−/−) had correspondingly longer period lengths. Mice that were behaviorally arrhythmic (Per2brdm/brdm, Per2brdm/brdm; Cry1−/−, and Per1brdm/brdm;Per2brdm/brdm) produced arrhythmic fibroblasts. In most cases, however, the period of fibroblast gene expression was more extreme than that of behavior. For example, mice with behavioral periods shorter or longer than wild-type gave fibroblasts whose periods were even shorter or longer still. Similarly, mice containing the double disruption Per2brdm/brdm;Cry2−/−, which are behaviorally rhythmic and have a period of 24.4 h, yielded fibroblasts that typically show faint rhythmicity of 23–29 h for one cycle before becoming arrhythmic (Figures 4 and 5).
Figure 4 Comparison of the Period of Wheel-Running Behavior and of Fibroblast Bioluminescence among Mouse Strains Containing Different Circadian Gene Disruptions
Mice of nine nearly isogenic genotypes were analyzed to obtain the period of running-wheel behavior (τrw) and the period of fibroblast luminescence (τf). For each genotype, a single running-wheel profile of the behavior of a single mouse kept first in normal light/dark conditions and then in constant darkness is shown. Periods of darkness are shaded on the graphs. Standard double-plotted actogram format is used, with consecutive rows representing consecutive days of activity, and x-axis showing time. Period lengths are presented below the graphs as average plus or minus the standard deviation for two mice, each measured twice. For each genotype, a single 100-h representative fibroblast recording is also shown. The x-axis shows time in hours; the y-axis shows arbitrary light units. Period lengths are presented above the graphs as averageplus or minus the standard deviation for two measurements of two biopsies of each mouse.
Figure 5 Genetic Differences in Circadian Period Length Measured from Fibroblasts Are Larger than Those Measured from Animal Behavior
Data from Figure 4 depicted in stacked bar graph format. In each rhythmic strain measured, the period of wheel-running activity is shown in light grey, expressed in the difference in hours from the 24-h solar day. On top of this is shown the change in period of fibroblasts from the same animals, also measured in the difference in hours from the solar day. Genotypes depicted, from left to right, are Per2brdm/brdm, Per1brdm/brdm, wild-type, Cry2+/−, Cry2−/−;Per1brdm/brdm, Cry2−/−, Per2brdm/brdm;Cry2−/−. Because Per2brdm/brdm;Cry2−/− mice had unstable periods that ranged widely from individual to individual, this genotype is shown twice at the extreme right, with representative mice with periods both less than and greater than 24 h separated into two groups. Arrhythmic fibroblasts are designated “arr.”
Table 1 Mouse Strains for Whom the Circadian Period of Running-Wheel Behavior Was Compared with the Period of Fibroblast Bioluminescence
Discussion
From the studies presented in this paper, we can conclude that molecular circadian rhythms can be measured in fibroblasts from skin biopsies and that the period of these rhythms is specific to an individual and can vary with genotype. Moreover, the variations observed among the 19 human individuals of this study suggest that the circadian clock is quite heterogeneous at a genetic level. The genetic origins of this variation will doubtless be a topic of future investigations, and our results suggest that fibroblasts could be an excellent system in which to investigate such differences by quantitative trait (QTL) mapping.
A major question posed by the research that we have presented is the relationship between fibroblast period length in vitro and the period length of human circadian physiology. Certainly, the two values are contingent upon the clocks of different tissues studied in different contexts (skin versus suprachiasmatic nucleus and in vitro versus in vivo). Although fibroblasts and SCN neurons possess clocks of very similar molecular mechanism [14], different mouse tissues from the same mouse can have periods varying by almost two hours when measured in tissue slices in vitro [8]. The average values that we obtained for the period of human circadian gene expression (24.5 h) correspond well with what has been published about rhythms of human circadian physiology (24.2–24.5 h) [21–24]. Values for period length in skin and fibroblasts of wild-type inbred mouse strains (23.5 ± 0.3 h) also corresponded nicely with behavioral and SCN period values obtained by us or published by others [8].
Nevertheless, our data suggest that it would be an error to assume that fibroblast period length is the same as physiologic period. Although on average the two corresponded well, significant differences were visible on an individual level. In mice, mutations at circadian loci that affected periodicity invariably had more extreme phenotypes upon fibroblast period than upon wheel-running period (Figures 4 and 5). Mice with behavioral periods 1 h shorter than wild-type gave fibroblasts whose period was 4 h shorter, and mice with periods 0.75 h longer had fibroblast periods 1.5 h longer. In most cases, though, longer or shorter periods of running-wheel behavior translate to longer or shorter fibroblast periods, respectively.
Our data hint at a similar disparity in humans. Specifically, the period lengths of human fibroblast gene expression showed inter-individual variation that was greater than what might have been expected from behavioral studies. Among our 19 human samples, a maximum difference of 4 h was seen, and six samples could be placed in categories that differ by more than 1.5 h. Altogether, a standard deviation of 0.8 h was observed. The period of circadian physiology measured by others in human beings showed standard deviations of 0.2–0.5 h under conditions of “forced desynchrony” during which circadian period was assessed in the absence of the confounding effects of light input and sleep-wake cycle feedback [21–24].
Although fibroblast clocks are not identical to SCN clocks, the fact that they use the same molecular components and that mutations at circadian loci affect biological timing in both tissues in the same qualitative fashion will doubtless render them quite useful in uncovering genetically-caused circadian differences among individuals and populations. Mammals other than humans can also show preferences of morningness and eveningness [25], and mouse wheel-running behavior has already been used as the basis for genome-wide quantitative trait analysis. In some human populations, chronotype questionnaires have suggested that morningness–eveningness tendencies can be widely distributed [26], and twin studies suggest that morningness and eveningness can be genetically determined [27]. Given its robustness, the presented procedure could be used in quantitative trait mapping of human period length and thus in the identification of genetic loci that participate in determining the period length and the phase of daily human rhythmicity. Ideally, studies with large numbers of human subjects should be performed with the least invasive cell harvesting techniques possible. Although the 2-mm cutaneous punch biopsies can be performed rapidly and heal completely within a few days, plucking hairs is obviously even less invasive. As shown in this paper, it was possible on occasion to harvest and cultivate primary keratinocytes from plucked hairs for the analysis of circadian gene expression. We hope that future efforts in optimizing this method will render it generally applicable.
Materials and Methods
Vector production
Figure 1A, construct (i): The EF1α promoter and gfp gene were removed from lentiviral backbone plasmid pWPI (http://www.tronolab.unige.ch, and replaced with a reporter cassette consisting of 1 kb of mouse Bmal1 upstream region and 53 nucleotides of exon 1, fused in-frame to the luciferase coding region, and followed by 1 kb of Bmal1 3′UTR. Two chicken β-globin FII elements [20] were synthesized by PCR and inserted on either side of the reporter cassette. Construct (ii): The Bmal1:luc reporter cassette was inserted downstream of an EF1α promoter and SV40 terminator in pWPI. All viruses were produced, concentrated 10-fold by ultracentrifugation, and used for infection as described [28].
Tissue isolation and culture
To establish our technique, five cylindrical 2-mm diameter cutaneous biopsies were taken from the ventral regions of five patients undergoing abdominoplasty operations. Subsequently, two biopsies were taken from the buttocks of each of seven recruited healthy adult subjects. Fibroblasts were isolated from biopsies by overnight digestion of tissue in DMEM/20% FCS/1 mg/ml collagenase type IA, and cultured in DMEM/20% FCS. Four foreskin fibroblast cultures were obtained by similar methods, and three other adult dermal fibroblast cultures were obtained from others (generous gift of S. Clarkson). Adult mouse fibroblasts from wild-type, Per1brdm/brdm, and Per1brdm/brdm; Per2brdm/brdm mice were isolated from 2-mm tail biopsies by the same method. Monocytes were isolated as described from fresh human blood from the Geneva University Hospital Blood Bank.[29]. Prior ethical consent for the use of all human tissues was given by the ethical committee of the Geneva University Hospital, informed consent was obtained from all human subjects, and animals were handled according to institutional guidelines.
Synchronization and measurement of circadian rhythms
Four days or more after infection or cell passage, circadian rhythms were synchronized by dexamethasone [18]. Medium without phenol red was supplemented with 0.1 mM luciferin, and light output was measured in homemade light-tight atmosphere-controlled boxes for at least 4 d [9].
Mouse running-wheel behavior
Mice of various genotypes were housed in cages with controlled lighting, each equipped with a running wheel (Mini Mitter, Bend, Oregon, United States) Running-wheel actograms and period determination were done with the Stanford Chronobiology kit (Stanford Software Systems, Stanford, California, United States).
Statistical methods
For each luciferase measurement, the period of oscillation was calculated by fitting the curve to sine waves of known period using a macro program for Microsoft Excel written by SAB. The maxima and minima of each oscillation were identified, and the timing of these points was used to fit hypothetical sine curves with period and phase as free variables. The period of the sine wave with the best least-squares fit to the data was assumed to be the true period of oscillation. Because the period length of the first day after synchronization varied according to the conditions of synchronization, it was not included in these calculations; rather, period was determined by analyzing only days 2–5. To determine the period length of a particular biopsy, two independent viral infections were performed, and two synchronization/measurement cycles were done for each infection. To determine the period length of a particular individual, at least two separate biopsies were analyzed in this manner. Values are presented as mean plus or minus the standard deviation.
Supporting Information
Accession Numbers
The Genbank (http://www.ncbi.nlm.nih.gov/Genbank) accession numbers for the genes discussed in this paper are chick β-globin (V00409), human CK1ɛ (L37043), human EF1α (X03558), human PER2 (AB002345), human PER3 (AB047686), mouse Bmal1 (Q9WTL8), mouse Cry2 (AB003433), mouse Per1 (AF022992), and mouse Per2 (AF036893).
The authors would like to thank Andre Liani, Yves-Alain Poget, and the Atelier of the University of Geneva for invaluable aid in constructing the real-time luminometer apparatus; S. Clarkson for the kind donation of primary fibroblast cultures; M. Wiznerowicz and D. Trono for furnishing lentiviral plasmids and expertise; and P. Paruch, C. Henchoz, C. Tschopp, A. Ruetschi, N. Reyren, M. Elmer, and F. Mauro for skin biopsies. Support for this study was provided by grants from the Swiss National Science Foundation, the Fondation Louis Jeantet de Medecine, and the National Centres of Competence in Research (NCCR) Frontiers in Genetics program.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. SAB conceived and designed the experiments. JMD and CAM supervised handling of human subjects. SAB, FFO, and CH performed the experiments. SAB and US analyzed the data. EN, CJ, JMD, RC, CAM, and UA contributed reagents/materials/analysis tools. SAB and US wrote the paper.
Citation: Brown SA, Fleury-Olela F, Nagoshi E, Hauser C, Juge C, et al. (2005) The period length of fibroblast circadian gene expression varies widely among human individuals. PLoS Biol 3(10): e338.
Note Added in Proof
The genome-wide quantitative trait analysis of mouse wheel-running behavior to which we refer in the Discussion was performed by Shimomura et al.: Shimomura K, Low-Zeddies SS, King DP, Steeves TD, Whitely A, et al. (2001) Genome-wide epistatic interaction analysis reveals complex genetic determinants of circadian behavior in mice. Genome Res 11: 959–980.
Abbreviations
LTRlong terminal repeat
SCNsuprachiasmatic nucleus
==== Refs
References
Schibler U Sassone-Corsi P A web of circadian pacemakers Cell 2002 111 919 922 12507418
Duffy JF Dijk DJ Hall EF Czeisler CA Relationship of endogenous circadian melatonin and temperature rhythms to self-reported preference for morning or evening activity in young and older people J Investig Med 1999 47 141 150
Duffy JF Rimmer DW Czeisler CA Association of intrinsic circadian period with morningness-eveningness, usual wake time, and circadian phase Behav Neurosci 2001 115 895 899 11508728
Toh KL Jones CR He Y Eide EJ Hinz WA An hPer2 phosphorylation site mutation in familial advanced sleep-phase syndrome Science 2001 291 1040 1043 11232563
Ebisawa T Uchiyama M Kajimura N Mishima K Kamei Y Association of structural polymorphisms in the human period3 gene with delayed sleep phase syndrome EMBO Rep 2001 2 342 346 11306557
Xu Y Padiath QS Shapiro RE Jones CR Wu SC Functional consequences of a CKIdelta mutation causing familial advanced sleep phase syndrome Nature 2005 434 640 644 15800623
Archer SN Robilliard DL Skene DJ Smits M Williams A A length polymorphism in the circadian clock gene Per3 is linked to delayed sleep phase syndrome and extreme diurnal preference Sleep 2003 26 413 415 12841365
Yoo SH Yamazaki S Lowrey PL Shimomura K Ko CH PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues Proc Natl Acad Sci U S A 2004 101 5339 5346 14963227
Yamazaki S Numano R Abe M Hida A Takahashi R Resetting central and peripheral circadian oscillators in transgenic rats Science 2000 288 682 685 10784453
Nagoshi E Saini C Bauer C Laroche T Naef F Circadian gene expression in individual fibroblasts: Cell-autonomous and self-sustained oscillators pass time to daughter cells Cell 2004 119 693 705 15550250
Welsh DK Yoo SH Liu AC Takahashi JS Kay SA Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression Curr Biol 2004 14 2289 2295 15620658
Welsh DK Logothetis DE Meister M Reppert SM Individual neurons dissociated from rat suprachiasmatic nucleus express independently phased circadian firing rhythms Neuron 1995 14 697 706 7718233
Liu C Weaver DR Strogatz SH Reppert SM Cellular construction of a circadian clock: Period determination in the suprachiasmatic nuclei Cell 1997 91 855 860 9413994
Yagita K Tamanini F van der Horst GT Okamura H Molecular mechanisms of the biological clock in cultured fibroblasts Science 2001 292 278 281 11303101
Pando MP Morse D Cermakian N Sassone-Corsi P Phenotypic rescue of a peripheral clock genetic defect via SCN hierarchical dominance Cell 2002 110 107 117 12151001
Tischkau SA Mitchell JW Tyan SH Buchanan GF Gillette MU Ca2+/cAMP response element-binding protein (CREB)-dependent activation of Per1 is required for light-induced signaling in the suprachiasmatic nucleus circadian clock J Biol Chem 2003 278 718 723 12409294
Nguyen TH Oberholzer J Birraux J Majno P Morel P Highly efficient lentiviral vector-mediated transduction of nondividing, fully reimplantable primary hepatocytes Mol Ther 2002 6 199 209 12161186
Balsalobre A Brown SA Marcacci L Tronche F Kellendonk C Resetting of circadian time in peripheral tissues by glucocorticoid signaling Science 2000 289 2344 2347 11009419
Schroder AR Shinn P Chen H Berry C Ecker JR HIV-1 integration in the human genome favors active genes and local hotspots Cell 2002 110 521 529 12202041
Bell AC West AG Felsenfeld G The protein CTCF is required for the enhancer blocking activity of vertebrate insulators Cell 1999 98 387 396 10458613
Czeisler CA Duffy JF Shanahan TL Brown EN Mitchell JF Stability, precision, and near-24-hour period of the human circadian pacemaker Science 1999 284 2177 2181 10381883
Kelly TL Neri DF Grill JT Ryman D Hunt PD Nonentrained circadian rhythms of melatonin in submariners scheduled to an 18-hour day J Biol Rhythms 1999 14 190 196 10452330
Carskadon MA Labyak SE Acebo C Seifer R Intrinsic circadian period of adolescent humans measured in conditions of forced desynchrony Neurosci Lett 1999 260 129 132 10025716
Robilliard DL Archer SN Arendt J Lockley SW Hack LM The 3111 Clock gene polymorphism is not associated with sleep and circadian rhythmicity in phenotypically characterized human subjects J Sleep Res 2002 11 305 312 12464098
Labyak SE Lee TM Goel N Rhythm chronotypes in a diurnal rodent, Octodon degus
Am J Physiol 1997 273 R1058 1066 9321886
Roenneberg T Wirz-Justice A Merrow M Life between clocks: Daily temporal patterns of human chronotypes J Biol Rhythms 2003 18 80 90 12568247
Drennan SJ Kripke DF Kelsoe J Gillin JC Morningness/eveningness is heritable Soc Neurosci Abstr 1992 196
Cepko C Ausubel FM Brent R Kingston R Moore DD Seidman JG Large-scale preparation and concentration of retroviral stocks Current protocols in molecular biology, volume 3 2001 New York John Wiley & Sons Chapter 16.21
Meier CA Chicheportiche R Dreyer M Dayer JM IP-10, but not RANTES, is upregulated by leptin in monocytic cells Cytokine 2003 21 43 47 12668159
Oster H Yasui A van der Horst GT Albrecht U Disruption of mCry2 restores circadian rhythmicity in mPer2 mutant mice Genes Dev 2002 16 2633 2638 12381662
Zheng B Albrecht U Kaasik K Sage M Lu W Nonredundant roles of the mPer1 and mPer2 genes in the mammalian circadian clock Cell 2001 105 683 694 11389837
Zheng B Larkin DW Albrecht U Sun ZS Sage M The mPer2 gene encodes a functional component of the mammalian circadian clock Nature 1999 400 169 173 10408444
van der Horst GT Muijtjens M Kobayashi K Takano R Kanno S Mammalian Cry1 and Cry2 are essential for maintenance of circadian rhythms Nature 1999 398 627 630 10217146
Oster H Baeriswyl S van der Horst GT Albrecht U Loss of circadian rhythmicity in aging mPer1-/-mCry2-/- mutant mice Genes Dev 2003 17 1366 1379 12782655
|
16167846
|
PMC1233413
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e338
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030338
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1617140810.1371/journal.pbio.0030342Research ArticleNeuroscienceCatCortical Sensitivity to Visual Features in Natural Scenes Cortical Feature SensitivityFelsen Gidon
1
2
¤Touryan Jon
2
3
¤Han Feng
2
3
Dan Yang [email protected]
1
2
3
1 Division of Neurobiology, Department of Molecular and Cell Biology, University of California, Berkeley, California, United States of America,2 Helen Wills Neuroscience Institute, University of California, Berkeley, California, United States of America,3 Group in Vision Science, University of California, Berkeley, California, United States of AmericaBurr David C. Academic EditorIstituto di NeurofisiologiaItaly10 2005 27 9 2005 27 9 2005 3 10 e34215 4 2005 3 8 2005 Copyright: © 2005 Felsen 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.
Complex Cells in the Brain's Vision Center Tune in to Natural Scenes
A central hypothesis concerning sensory processing is that the neuronal circuits are specifically adapted to represent natural stimuli efficiently. Here we show a novel effect in cortical coding of natural images. Using spike-triggered average or spike-triggered covariance analyses, we first identified the visual features selectively represented by each cortical neuron from its responses to natural images. We then measured the neuronal sensitivity to these features when they were present in either natural images or random stimuli. We found that in the responses of complex cells, but not of simple cells, the sensitivity was markedly higher for natural images than for random stimuli. Such elevated sensitivity leads to increased detectability of the visual features and thus an improved cortical representation of natural scenes. Interestingly, this effect is due not to the spatial power spectra of natural images, but to their phase regularities. These results point to a distinct visual-coding strategy that is mediated by contextual modulation of cortical responses tuned to the spatial-phase structure of natural scenes.
By recording how visual cortical neurons respond to natural scenes versus random stimuli, the authors discover that complex cells, but not simple cells, respond more sensitively to the phase regularities of natural images
==== Body
Introduction
An essential goal in studying the visual system is to understand how it processes natural scenes, which exhibit distinct statistical properties [1–7]. In particular, since neuronal circuits evolve and develop in the natural environment, they may be specifically adapted for efficient coding of natural stimuli [1–4,8–10]. This efficient coding hypothesis has provided an important framework for understanding early visual processing: In the retina and the lateral geniculate nucleus, spatiotemporal frequency tuning of the neurons appears to be adapted to the power spectra of natural scenes, allowing them to encode the input into a more efficient, decorrelated form [3,11–13]. In the primary visual cortex (V1), response properties of both simple [14–16] and complex [17,18] cells can be “derived” from the statistics of natural scenes and the principle of efficient coding. These studies underscore the importance of understanding neuronal response properties with respect to natural-scene statistics.
The function of neurons in the early visual pathway is thought to be the analysis of local features in the images. The response of each neuron is determined by two properties: the structure of its preferred visual features and the sensitivity of the neuron to the presence of these features in visual scenes. The preferred features are closely related to the receptive field (RF) of the cell. For a neuron with linear RF properties (e.g., a simple cell), the preferred feature directly corresponds to the classical RF (i.e., light and dark regions correspond to ON and OFF response regions). In a standard model of the complex-cell RF, there are multiple preferred features, and each feature corresponds to the RF of a functional subunit [19–21] or a linear combination of them. Feature sensitivity, on the other hand, can be characterized by the contrast-response function, which describes the relationship between the neuronal response and the contrast of the feature (see below). High feature sensitivity allows the neuron to reliably signal the presence of the preferred features in visual stimuli. Note that, in this study, we use the term “preferred feature” solely to facilitate the discussion of our findings on cortical feature sensitivity; we do not imply that these features are necessarily optimal for driving the cortical neuron, as the neuronal responses are known to be modulated by various contextual stimuli, including those in the nonclassical RF [22].
The structure of the preferred features of V1 neurons has been investigated extensively in experimental studies [19,20,23,24], and is thought to play an important role in efficient, sparse coding of natural scenes [14–18]. However, the role of feature sensitivity in efficient coding has not been studied experimentally. In the present study, we measured feature sensitivity of cortical neurons in processing several classes of visual stimuli. By comparing the cortical feature sensitivity for natural images, random stimuli, and synthetic stimuli with either natural power or natural phase spectra, we characterized the dependence of feature sensitivity on the statistical properties of visual stimuli. Our results point to a novel form of efficient coding that is mediated by contextual modulation of cortical responses.
Results
Measurement of Preferred Features and Feature Sensitivity
Single-unit recordings were made from 50 neurons (36 complex cells, 14 simple cells) in area 17 of anesthetized adult cat (see Materials and Methods). We will first report the findings from complex cells and will then describe the results from simple cells in a separate section. In order to characterize the feature sensitivity of each complex cell, it is necessary to first estimate its preferred features. The preferred features were estimated from the responses of the neuron to a natural-image ensemble (Figure 1A), using a spike-triggered covariance (STC) analysis [25–27] (see Materials and Methods). These features, which are represented by the “significant eigenvectors” of the STC matrix, contained oriented light and dark regions (Figure 1B), resembling the RFs of simple cells [19]. For the majority of the cells analyzed with this method (17/26), we identified two significant eigenvectors, which are well approximated by a pair of Gabor functions with similar orientations and spatial frequencies but ∼90° phase difference (Figure 1B, left panel). For other cells, only one significant eigenvector was identified, which was also well approximated by a Gabor function.
Figure 1 Measurement of Preferred Features and Feature Sensitivity for V1 Complex Cells
(A) Upper panel: example natural images. White boxes (12 × 12 pixels) indicate area presented in experiments. Lower panel: schematic spike train, binned at stimulus frame rate (24 Hz, dotted lines). Arrow indicates temporal delay (1 frame) at which preferred features were estimated, which was determined in preliminary studies to be the optimal temporal delay (see Figure S2).
(B) Estimation of preferred features (significant eigenvectors) using STC analysis (see Materials and Methods). Left panel: preferred features of a neuron, with light and dark regions represented by red and blue; dashed ovals delineate the first feature to facilitate comparison with the images. Right panel: 30 largest eigenvalues of STC matrix. Dashed lines: control confidence intervals (mean ± 12 standard deviation of control eigenvalues). Filled circles: significant eigenvalues corresponding to eigenvectors shown on left.
(C) Upper panel: natural images. Dashed ovals correspond to those in (B). Middle panel: contrast of the first preferred feature (F.C. denotes feature contrast; see Materials and Methods). Lower panel: responses of the neuron (in spikes/s) to natural images. Black dots: feature contrasts (middle) and neuronal responses (lower) for the example images.
(D) Contrast-response function. Error bar: ± standard error of the mean.
The contrast of each significant eigenvector in a particular image (referred to as “feature contrast”) was measured as the dot product of the image and the eigenvector (Figure 1C), which depends on both the overall contrast of the image and its similarity with the eigenvector (e.g., the image on the right in Figure 1C contains a high-contrast luminance edge matched to the first eigenvector, thus giving rise to a high feature contrast). The sensitivity of the neuron to each of its significant eigenvectors in a stimulus ensemble is measured by plotting the average neuronal response as a function of feature contrast (Figure 1D), yielding the contrast-response function [26,28,29] (see Materials and Methods). A steep contrast-response function indicates a high sensitivity of the neuron to the presence of the corresponding feature in the images, whereas a flat contrast-response function indicates that the neuronal response is insensitive to the presence of the feature.
The structure of the significant eigenvectors (Figure 1B) and the shape of their contrast-response functions (Figure 1D) are qualitatively consistent with the standard model of complex-cell RFs [19–21]. This “energy model” can be described as
where r is the response of the neuron, S is the stimulus, k
φ is the RF of a subunit with preferred spatial phase φ (k
φ, k
φ+90°, and their linear combinations also correspond to the preferred features of the cell), and F represents the contrast-response function for the subunit, which is commonly approximated by F(x) = βx
2. Given the quadratic nonlinearity in this model [21], the STC analysis provides an ideal method for estimating the preferred features of each complex cell: The RF of each subunit (k
φ or k
φ+90°) corresponds to a significant eigenvector (Figure 1B) or a linear combination of the two eigenvectors [26], and the contrast-response function F(x) can be measured as shown in Figure 1C and 1D. It is important to note, however, that the STC analysis does not presume the validity of the energy model [30]; conversely, the results of the analysis (Figure 1B and 1D) do not imply that the energy model provides a complete description of complex-cell responses, as shown below.
Feature Sensitivity of Complex Cells in Response to Natural and Random Stimuli
According to the energy model, the response of a complex cell to each image is completely determined by the contrasts of the preferred features in the image, k
φ•S and k
φ+90°•S. However, numerous studies have shown that many visual stimuli that are ineffective in driving a cortical neuron on their own (e.g., stimuli at non-preferred orientations or in the nonclassical RF) may strongly modulate the responses to the effective stimuli [22,31,32] and affect the neuronal sensitivity to the preferred features [26,30]. To test whether cortical feature sensitivity depends on the statistics of the images that contain both the preferred and non-preferred features, we measured the responses of complex cells to natural images and to spatially unstructured random stimuli (Figure 2A). To control for the confounding effect of contrast adaptation [33] on cortical feature sensitivity, we constructed the random-stimulus ensemble specifically for each cell such that it matched the natural-image ensemble, frame by frame, in both the global (root-mean-square) contrast and the contrasts of one or both of the preferred features of the cell (Figure 2). This was achieved by first creating an orthonormal basis set that included the significant eigenvector(s) of the cell. Then, for each natural image, a random stimulus was generated by selecting the coefficients of the basis functions such that the global and feature contrasts were matched between the natural and random stimuli (see Materials and Methods). Because of the randomness in generating most of the basis functions and in selecting their coefficients, these stimuli exhibit no clear spatial structure (Figure 2A).
Figure 2 Matching of Feature Contrasts in Natural and Random Ensembles
(A) Example images in the natural (upper row) and the random (lower row) ensembles, which were matched frame by frame for both global and feature contrasts.
(B) Contrasts of a preferred feature of a complex cell (inset at center) in each frame of the natural (squares) and random (circles) ensembles in (A). F.C. denotes feature contrast.
(C) Distributions of feature contrasts in the natural (left) and random (middle) ensembles, and the distribution of the difference in feature contrast between the two ensembles (right).
When we examined the contrast-response functions that were computed, for each significant eigenvector, from the responses to the natural- and random-stimulus ensembles, we found that, for both ensembles, the neuronal response increased with the magnitude of the feature contrast independently of its sign (Figure 3A), consistent with the polarity invariance of complex cells [19,20]. Interestingly, however, the amplitude of the contrast-response function was markedly higher for natural images than for random stimuli. To compare these functions quantitatively, we fit each function with
Figure 3 Feature Sensitivity of Complex Cells in Response to Natural Images and Random Stimuli
(A) Contrast-response functions for both preferred features (insets above) of a complex cell. Curves: fits of data with quadratic functions.
(B) Gain of contrast-response function (in spikes/s per unit feature contrast) for natural ensemble versus that for contrast-matched random ensemble. For this population of cells, the gain was significantly higher for the natural than for the random ensemble (n = 24, from 14 cells; p < 10−4, Wilcoxon signed rank test).
where r is the neuronal response, x is feature contrast, γ = 2 (based on the standard energy model of complex-cell RFs [21]; see above), and r
0 and β are free parameters (Figure 3A, curves). For all the cells tested in this experiment (14 cells, 24 significant eigenvectors), we found that the gain of the contrast-response function (β), which directly reflects the feature sensitivity of the neuron, was higher in the responses to natural images (p < 10−4, Wilcoxon signed rank test; Figure 3B). This result was independent of the value of γ between one and three, and it remained the same if the positive and negative sides of each function were fitted separately.
In the above experiment, ten of the 14 complex cells had two significant eigenvectors. For four of these ten cells, the natural and random stimuli were matched for the feature contrasts of both eigenvectors simultaneously. For the remaining six cells, however, a random-stimulus ensemble was generated to match the natural ensemble for each eigenvector separately. In this case, it is possible that the contrasts of the two significant eigenvectors were more correlated in natural images than in the random stimuli [34], and the response of the cell to the matched eigenvector may be enhanced by the correlated presence of the other eigenvector. To test this possibility, we re-computed the contrast-response function for each significant eigenvector of the ten cells using only the frames in which the contrast of the other “un-matched” eigenvector was very low (<0.005), so that excitation of the cell due to this un-matched feature was negligible (Figure 3A). We found that the feature sensitivity was still significantly higher for the natural than for the random stimuli (p < 0.002, Wilcoxon signed rank test). Furthermore, if we consider only the four cells for which the contrasts of both significant eigenvectors were matched simultaneously (thus any correlation between them in natural images would be equally present in the random stimuli), the feature sensitivity was still significantly higher for the natural than for the random stimuli (p < 0.01). Thus, the higher feature sensitivity of the complex cells in response to natural images is not due to any correlation between the two significant eigenvectors.
Another potential problem in the above experiment is that the preferred features were estimated from the responses to natural images, the non-Gaussian statistics of which may cause bias in the estimation of the preferred features with STC. In addition, although for most of the cells (12/14) the preferred features and their contrast-response functions were computed from the responses to different repeats of the natural-image ensemble, it is possible that using the same stimuli for both computations can introduce bias in the measured sensitivity. To control for these potential biases, for a separate set of complex cells we replaced the features estimated with natural images (see Figure 1B) with Gabor functions whose parameters were chosen based on the orientation and spatial-frequency tuning of the neuron measured with sinusoidal gratings. These Gabor functions resembled the significant eigenvectors measured with natural images (Figure 1B). When we compared the responses to a natural-image ensemble and a random-stimulus ensemble matched for both the global contrast and the contrast of the Gabor function (analogous to the analyses shown in Figures 2 and 3), the neuronal sensitivity to these Gabor functions was also found to be higher in the responses to natural images (n = 10, from ten cells; p < 0.05). This indicates that the effect is robust with respect to small variations in the spatial structure of the preferred visual features, and it is not due to the potential biases in feature estimation with natural images. Such an effect is not predicted by the existing models of complex-cell RFs [21,34,35], and it indicates that cortical feature sensitivity depends strongly on the image statistics.
Feature Sensitivity of Simple Cells
Surprisingly, for simple cells we found no difference in feature sensitivity between their responses to natural and random stimuli. The preferred feature of each simple cell (Figure 4A, inset), which directly corresponds to the classical RF of the cell, was estimated with a modified spike-triggered average (STA) analysis that corrects for the spatial correlations in natural images [23,36] (see Materials and Methods). We then constructed a random-stimulus ensemble that was matched to a natural-image ensemble, frame by frame, for both the global contrast and the contrast of the preferred feature (similar to the method for complex cells; see Figure 2). Responses to these contrast-matched natural and random ensembles were recorded, and the contrast-response functions were computed. Not surprisingly, the contrast-response functions measured with both ensembles were monotonic (see Figure 4A), consistent with the polarity sensitivity of simple cells [19]. To obtain a quantitative measure of the simple-cell feature sensitivity, we fit the positive side of each contrast-response function with
Figure 4 Feature Sensitivity of Simple Cells in Response to Natural Images and Random Stimuli
(A) Contrast-response function for the preferred feature (inset above) of a simple cell. Curves: fits of data with quadratic functions (for positive feature contrasts only).
(B) Gain of contrast-response function, as in Figure 3B (n = 14, from 14 cells).
where r is the neuronal response, x is the feature contrast, γ = 2, and r
0 and β are free parameters. Across the population of simple cells studied, the contrast-response gain (β) was not significantly different between the natural images and the random stimuli (n = 14, from 14 cells; p > 0.55) (Figure 4B). To compare the results between simple and complex cells directly, for each cell we computed Δβ = βnatural − βrandom (βnatural and βrandom are contrast-response gains for natural and random stimuli, respectively; for cells with two significant eigenvectors, Δβ was averaged between the two). We found that Δβ is significantly higher for complex cells than for simple cells (p < 0.001, Wilcoxon rank sum test).
Note that, in this study, a cell was classified as simple if F
1
/F
0 > 0.6 (see Materials and Methods). This criterion is somewhat arbitrary, as V1 neurons may lie on a simple-complex continuum rather than belong to one of two distinct categories [37–39]. To test whether our classification criterion affects the observed difference in feature sensitivity between simple and complex cells, we plotted Δβ against F
1/F
0 for each cell (Figure 5). While the distributions of both Δβ and F
1/F
0 may be continuous, there is a significant negative correlation between Δβ and F
1/F
0 (p < 10−4), and the cells with the highest F
1/F
0 (most “simple-cell like”) tend to exhibit the lowest Δβ. If we use the standard criterion [40] and classify the five cells with 0.6 < F
1/F
0 < 1 as complex cells (thus using the first eigenvector of their STC instead of STA as the preferred feature), Δβ is still significant for complex cells (p < 5 × 10−4, Wilcoxon signed rank test) but not for simple cells (p > 0.4), and Δβ is significantly different between simple and complex cells (p < 10−4, Wilcoxon rank sum test). Thus, the observed difference in feature sensitivity between simple and complex cells is not sensitive to the criterion used for simple/complex classification. Additional analyses have demonstrated that this difference is also not due to the different methods (STA versus STC) used to identify the preferred visual features (Protocol S1; Figure S1).
Figure 5 Difference in Feature Sensitivity between the Responses to Natural and Random Stimuli as a Function of F
1
/F
0
Each symbol represents data from one cell. For complex cells with two significant eigenvectors, the sensitivity difference was averaged between the two eigenvectors. Dashed line: linear fit.
Detectability of Preferred Features from Complex-Cell Responses
Functionally, the observed difference in complex-cell feature sensitivity predicts that the preferred features are more detectable in natural images than in random stimuli, but this difference could be eliminated if the increased sensitivity is accompanied by a similar increase in response noise. We thus tested this prediction directly using signal detection theory [41]. For simplicity, we defined two sets of stimuli in each ensemble: those with high feature contrast (>T
1; Figure 6A, black shading, referred to as “feature present”) and those with near-zero feature contrast (<T
0; gray shading, “feature absent”). Detectability of the feature was measured by how reliably the “feature-present” stimuli could be distinguished from the “feature-absent” stimuli based on the neuronal response (see Materials and Methods). Figure 6B shows the probability distribution of the neuronal response when a feature was either present (solid line) or absent (dashed line), for both the natural- and random-stimulus ensembles. For the natural ensemble, the neuron was more likely to fire at higher rates when the feature was present than when it was absent, as expected. Such a difference in the response probability allows correct classification of the two sets of stimuli at a level well above chance (50%; see Figure 6C).
Figure 6 Detectability of Features from Neuronal Responses to Natural Images and Random Stimuli
(A) Probability distribution of feature contrast in a natural ensemble (or, equivalently, its matched random ensemble). For simplicity, only the positive side (feature contrast >0) is shown. Gray shading: feature contrasts near zero (<T
0, here T
0
= 0.007, “feature absent”); black shading: high feature contrasts (>T
1, here T
1 = 0.04, “feature present”).
(B) Conditional probability distributions of responses evoked by natural images (upper) and random stimuli (lower). Solid lines: response distributions when the feature was present in stimulus (black shading in [A]); dashed lines: distributions when the feature was absent (gray shading in [A]).
(C) Feature detectability in natural images versus that in matched random stimuli, for the same population of cells shown in Figure 3B. Detectability was measured as the percentage of trials in which stimuli were correctly classified as “feature present” or “feature absent” (see Materials and Methods).
For the random ensemble, however, the two response distributions were much less distinguishable (Figure 6B), resulting in a lower percentage of correct classifications. This result is inconsistent with the energy model, which would predict that the upper and lower plots in Figure 6B would be identical. For the population of cells studied, detectability of the preferred features from the neuronal responses was significantly higher for the natural-stimulus ensemble (n = 24, p < 10−4, Wilcoxon signed rank test; Figure 6C), and this result was independent of the criteria used to select the two stimulus sets (Figure 6A, T
0 and T
1).
Dependence of Cortical Feature Sensitivity on Power and Phase Spectra
What stimulus property is responsible for the higher feature sensitivity of complex cells in response to natural rather than to random stimuli? Since the natural- and random-stimulus ensembles were matched for both global and feature contrasts (see Figure 2), the difference in contrast-response gain (see Figure 3) cannot be attributed to cortical contrast adaptation [33]. Instead, it is likely due to the differences in the spatial characteristics of the stimuli. For natural images, power (P) decreases with spatial frequency [1,2], and nearby frequencies tend to have similar phases (φ) [5], due to the prevalence of surfaces and edges of objects. White noise, on the other hand, has a flat power spectrum and random phase structure. For convenience, we use P
+/φ+ and P−/φ− to represent the statistical properties of natural and white-noise stimuli, respectively (where “+” represents natural). In order to distinguish the effects of power and phase spectra on cortical feature sensitivity (Figure 3), we manipulated each property separately, yielding two classes of synthetic stimulus ensembles: the “natural-power” ensemble, in which each image had a natural power but a random phase spectrum (P
+/φ−), and the “natural-phase” ensemble, in which each image had a random power but a natural phase spectrum (P−/φ+) (Figure 7A; see Materials and Methods).
Figure 7 Effects of Power and Phase Spectra of Stimuli on Cortical Feature Sensitivity
(A) Four classes of stimulus ensembles with distinct combinations of power (P) and phase (φ) characteristics; +: natural; −: random. Example stimuli from each class are shown. The P−/φ− and P−/φ+ stimuli are matched for both the global contrast and the feature contrasts for a particular complex cell.
(B) Summary of cortical feature sensitivity (contrast-response gain; see Figure 3B) for the stimulus classes in (A). In each experiment, a random (P−/φ−) stimulus ensemble was generated to match P
+/φ+, P
+/φ−, or P−/φ+ in global and feature contrasts (see Figure 2 and Materials and Methods), and the measured contrast-response gain was plotted against the gain for P−/φ− (as in Figure 3B). Bar represents slope of linear regression (through origin); >1 indicates higher contrast-response gain relative to P−/φ−. Error bar: ± standard deviation. P
+/φ+ bars for simple (S) and complex (C) cells were computed from data in Figures 3B and 4B, respectively, and P
+/φ− (n = 10, from six cells) and P−/φ+ (n = 11, from six cells) were from largely nonoverlapping populations of complex cells (one cell was used in two separate experiments).
Visually, natural-power stimuli tend to exhibit smooth luminance variations in space (typical of natural images but not of white noise), but they appear amorphous owing to the lack of well-defined edges and contours. Natural-phase stimuli, however, contain reduced low-frequency signals but retain and enhance the edges in natural images. Thus, these synthetic stimuli capture complementary spatial characteristics of natural images.
When we compared the feature sensitivity of each complex cell in response to each of these ensembles with that to a random ensemble (matched for global contrast and the feature contrasts for all significant eigenvectors; see Materials and Methods), we found that the contrast-response gain was significantly higher for P−/φ+ (n = 11, from seven cells; p < 0.02, Wilcoxon signed rank test) but not for P
+/φ− (n = 10, from six cells; p > 0.15) (Figure 7B). This indicates that the increased feature sensitivity is due not to the power spectrum of natural images, but to their spatial-phase regularities.
Discussion
The main finding of this study is a new response property of complex cells pertaining to the coding of natural stimuli. Although the hypothesis of efficient coding by cortical neurons in terms of redundancy reduction among a population of neurons has been examined in theoretical studies [2,4,9,10,14–18], there have been relatively few experimental studies on the relationship between cortical responses and the statistical properties of natural stimuli [24,42,43]. In the present study, we have shown that the response sensitivity of complex cells to their preferred features is higher for natural images than for random stimuli (see Figure 3), leading to increased feature detectability in the cortical representation of natural stimuli (see Figure 6).
Since these stimulus ensembles were matched frame by frame for the contrasts of the significant eigenvectors (k
φ•S and k
φ+90°•S; see Figure 2), they should activate the energy-model mechanism to the same extent. The observed difference in cortical responses can therefore be attributed to the numerous other features that were not matched between the two ensembles (see Materials and Methods). Previous studies have shown that the non-preferred stimuli can affect the gain of the neuronal contrast-response function for the preferred stimuli [26], and these effects (which we refer to as “contextual modulation”) have been modeled as divisive normalization [34,35]. This normalization model can be described as
where ∑φ
F
φ(k
φ•S) corresponds to the energy model, kj and Fj are the RF and contrast-response function of the jth divisive subunit, respectively, and σ is a constant. This model can account for a range of nonlinear cortical-response properties, including contrast-gain control and cross-orientation inhibition [44]. Our finding suggests that the non-preferred features (kj) provide different degrees of suppressive modulation of the responses to the natural and random stimuli, resulting in higher feature sensitivity for natural images. Note that while gain control and contextual modulation may enhance the efficiency of visual coding by optimizing information transmission in individual neurons [45], reducing inter-neuronal correlations [34,42], or increasing response sparseness [42,46], a potential detrimental effect is a reduction in the sensitivity of cortical neurons to their preferred features when the features are embedded in complex stimuli. Our results suggest that this mechanism is tuned to the spatial statistics of natural images so as to reduce the suppression of the responses to the preferred features.
A well-known form of suppressive modulation is cross-orientation inhibition [32,47], and the higher feature sensitivity of complex cells in response to natural images could be caused by fewer cross-oriented components in these stimuli. To test for this possibility, we estimated the contrast energy at the orientation orthogonal to the preferred features as the mean-square contrast of the preferred features rotated by 90°. We found no significant difference in this contrast energy between the natural- and the random-stimulus ensembles (p > 0.4). In addition to the total energy over the entire ensemble, we also examined whether, in each stimulus frame, the contrast energy of the preferred features is correlated with the energy of the cross-oriented components. We found that such correlation is negligible (correlation coefficient within ±0.01) in both the natural and random ensembles. These results suggest that the observed difference in the feature sensitivity of complex cells in response to the natural and random stimuli is not likely to be due to the difference in cross-orientation inhibition. We also considered the possibility that the natural images contain less global contrast (and thus less non-preferred feature contrast) than the random stimuli within the RF of the cell, since matching the stimuli for global contrast over the entire image does not guarantee a precise match within the RF region. However, further analyses suggested that this could not account for the observed difference in feature sensitivity (Protocol S2).
Aside from suppressive modulation, there may be additional excitatory visual features (k
φ) not identified by our STC analysis, a possibility indicated by a recent study in macaque V1 [30]. These additional excitatory features may be more correlated with the identified features in natural images than in random stimuli, resulting in higher response gains for the identified features. If this were the case, it would suggest that the improved feature sensitivity of complex cells is mediated by tuning of the additional excitatory features to the statistics of natural images. Note that, in the traditional view, the function of complex cells in detecting oriented edges or lines is primarily mediated by a pair of Gabor filters described by the energy model. Our current finding, together with that in macaque V1 [30], indicates that these cells have more specialized RF properties. They can carry out more refined feature detection by responding more vigorously to the oriented edges and contours that are meaningful features in natural images (e.g., those belonging to the borders of physical objects) than to random stimuli that contain the same contrasts of the pair of Gabor filters.
Interestingly, while this property is highly robust in complex cells (see Figure 3), it is virtually absent in simple cells (see Figures 4 and 5). This difference may be due to the differential laminar distributions of the two cell types and the different neuronal circuitry contributing to their contextual modulation. Related to our finding, an earlier study showed that the repulsive shift in cortical orientation tuning induced by surround visual stimulation, which may increase the efficiency of visual coding, is found only for complex cells and not for simple cells [48].
A study of feature detection in human vision led to the suggestion that complex cells act as detectors for phase congruence in visual images [49]. Recent physiological experiments have shown that the responses of complex cells to compound gratings depend on the relative spatial phase of the gratings [50]—an effect not accounted for by the energy model. In the present study, the higher feature sensitivity to natural images is also due to their nonrandom phase structure rather than to their power spectra (see Figure 7). This indicates that the phase regularity of natural images, due in large part to the prevalence of well-defined edges and contours [5,49], can strongly affect the cortical-response gain. This effect and that observed with compound gratings [50] may share common mechanisms. At the perceptual level, although the non-flat power spectrum is a well-studied, robust feature of natural scenes [1,2], the phase spectrum in fact carries most of the information that allows the animal to distinguish one scene from another [51–53]. Thus, along the mammalian visual pathway, coding of natural scenes appears to be refined at multiple stages: While the RF structure of retinal and thalamic neurons is adapted to the power spectra of natural stimuli to reduce coding redundancy [3,11–13], and the RFs of cortical simple cells are adapted to the phase structure of natural stimuli to attain sparse coding [14–16], gain control of feature sensitivity of complex cells is tuned to the phase regularity of natural scenes to improve the saliency of relevant visual features.
Materials and Methods
Overview of experimental paradigm
In this study, each cortical neuron was subjected to a sequence of inter-dependent experiments and analyses. To facilitate understanding of this experimental design, we provide in this section a brief outline of the major steps involved in studying each cell (details of these steps are provided in subsequent sections).
In Step 1, we estimated the preferred feature(s) of the cell from the recorded responses to an ensemble of natural images (see Figure 1A), using STA for simple cells and STC for complex cells.
In Step 2, we created a random-stimulus ensemble that was matched to a “target ensemble” (which was a natural-image, a natural-phase, or a natural-power ensemble; see Visual stimulation) for both global and feature contrasts (see, for example, Figure 2). Note that this step depends on the outcome of Step 1, since matching the feature contrast requires knowing the precise structure of the preferred feature(s).
In Step 3, we recorded the responses of the neuron to both the target and the random ensembles (created in Step 2) and computed the contrast-response functions for each feature from both responses (see Figure 3A).
In Step 4, we compared the gains (β) of the contrast-response functions computed from the responses to the target and random ensembles (see Figures 3B, 4B, and 7B).
In one control experiment, we replaced the preferred features estimated in Step 1 with Gabor functions whose parameters were chosen based on the responses of the cell to drifting gratings. Steps 2–4 then followed as above.
Recording
Animal-use procedures were as previously described [26] and approved by the Animal Care and Use Committee at the University of California, Berkeley. A total of 18 cats (weighing 2–6.5 kg each) were used. Single-unit recordings were made in area 17 using tungsten electrodes (A-M Systems, http://www.a-msystems.com). Cells were sampled at all laminar locations. Unit isolation was based on cluster analysis of waveforms. Cells were excluded if their mean firing rates were <1 spike/s or if their response correlations between repeats were at chance level. The firing rates of the cells included in this study ranged from 1 to 76 spikes/s (median: 6 spikes/s). Cells were classified as simple or complex based on the ratio of the first harmonic (F
1) and the mean (F
0) of the response to a drifting grating stimulus [40] (simple cell if F
1
/F
0 > 0.6; this criterion was used because for all the cells in our sample with 0.6 < F
1
/F
0 < 1, RFs measured by STA exhibited clear spatial structure).
Visual stimulation
Visual stimuli were generated with a PC and presented with a Barco monitor (size 40 × 30 cm, refresh rate 120 Hz, maximum luminance 80 cd/m2). Luminance nonlinearities were corrected through software. Four classes of stimulus ensembles were used in this study: natural-image, natural-power, natural-phase, and random-stimulus ensembles.
Natural-image ensemble (P
+/φ+)
Raw images were selected at random from a database consisting of a variety of digitized natural movies [16], and the center patch (12 × 12 pixels) of each image was retained. To maximize the diversity of images, we measured the similarity between each pair of stimuli Si and Sj in the ensemble by their dot product
For similar images (dot product > 0.95), either Si or Sj was excluded. Three distinct natural ensembles were used in this study (with no common image between ensembles). Unlike natural movies, these images were presented in a random sequence; the absence of temporal correlation greatly facilitated the STC analysis [27] (see below). However, since natural images are highly variable in their global contrast (measured by root-mean-square contrast
where −1 ≤ S(x, y) ≤ 1, and 1 and −1 represent highest and lowest luminance of the monitor, respectively), such stimuli may invoke contrast adaptation [33] that could confound the interpretation of the results. To control for this problem, we scaled each image such that all frames had the same global contrast (0.32). Any frame that could not be scaled without violating −1 ≤ S(x, y) ≤ 1 was excluded from the ensemble.
Natural-power ensemble (P
+/φ−)
First, we computed the Fourier transform of every image in a natural ensemble (P
+/φ+). We then generated each image in the P
+/φ− ensemble as
where Vj is the jth Fourier component of the corresponding natural image, Aj is the amplitude spectrum of the natural image, and the phase spectrum (φj) was obtained from a randomly generated 12 × 12 image, in which the luminance value of each pixel was drawn randomly from a uniform distribution on [0, 1]. The resulting image had the same power spectrum as its corresponding natural image, but the spatial phase spectrum is random. The global contrast of each image was the same as that for the natural images (0.32).
Natural-phase ensemble (P
−/φ+)
For each image in the natural ensemble (P
+/φ+), we retained its phase spectrum, but drew the amplitude at each frequency randomly from a uniform distribution on [0, 1], yielding the corresponding image in the natural-phase ensemble. These natural-phase images largely preserve the edges in natural images, but the spatial power spectrum of each image is random; the average power spectrum over the entire ensemble is flat.
Random-stimulus ensemble (P
−/φ−)
We generated the random ensemble specifically for each cell and each target ensemble (which could be a natural-image, natural-power, or natural-phase ensemble) to be matched for both the global and feature contrasts. First, we created an orthonormal basis set (Vi, where i = 1, 2, …, 144) that included the preferred feature(s) of the cell (i.e., for a complex cell with two significant eigenvectors that are matched simultaneously, V
1 and V
2 represent these eigenvectors, and V
3, … V
144 are arbitrary aside from the requirement of orthonormality; for a simple cell, V
1 represents its preferred feature measured by STA, and V
2, … V
144 are arbitrary). This was achieved in the following steps.
First, we generated a 144 × 144 symmetric matrix from the two significant eigenvectors (k
φ and k
φ+90°) as
where a
1 and a
2 were large but unequal numbers (e.g., a
1 = 109, a
2 = 108).
Second, we generated another 144 × 144 symmetric matrix using random vectors (Ui, where i = 1, 2, …, n, and where n >> 144) as
with each component of Ui drawn randomly from a normal distribution (mean = 0, variance = 1).
Third, we calculated the eigenvectors of Xvector + Xrandom (Vi, where i = 1, 2, …, 144), and these eigenvectors were used as the basis set for generating the random stimuli. Note that the large coefficients a
1 and a
2 ensure that the two preferred features (k
φ and k
φ+90°) can be arbitrarily close to the first two eigenvectors (V
1 and V
2).
Then, for each image in the target ensemble, we generated a corresponding “random” stimulus as
The coefficients ci were selected such that (1) c
1 and c
2 were the contrasts of the first and second preferred features in the target image, which ensured that the random stimulus was matched to the target image for feature contrast (see Figure 2), and (2) ci,where i = 3, 4, …, 144 were drawn randomly from a normal distribution and then scaled so that
where S is the target image. This ensured that the global stimulus contrast (equal to
for an orthonormal basis set) was also matched to that of the target image. If any pixel of Sr(x, y) fell outside the range [−1, 1], this frame was discarded and new ci were selected, until −1 ≤ S r(x, y) ≤ 1. This process was repeated for all the frames in the target ensemble. Note that although the stimuli that satisfied constraints (1) and (2) are, strictly speaking, not random, they exhibit no clear spatial structure (Figure 2A) because of the randomness in choosing most of the basis functions (V
3, …, V
144) and their coefficients (c
3, …, c
144). The spatial power spectra of these stimuli are flat, as for white noise.
All the stimuli described above were updated every five frames, corresponding to an effective frame rate of 24 Hz. Each ensemble consisted of 24,000 effective frames (referred to in the main text simply as “frames”) and was 16.7 min long. For each cell, the size of the images was adjusted to be slightly larger than the classical RF, although across the population of cells the relationship between the RF and the stimulus patch varied somewhat with respect to location and size. In experiments comparing feature sensitivity, we interleaved the presentations of target and random ensembles in order to control for the effects of the slow drift in the physiological state of the animal.
STC analysis
This technique has been used in previous studies to analyze the nonlinear-response properties of sensory neurons [25] and has been shown to be effective for computing the preferred features of V1 complex cells [26,27,30]. For all the cells in this study (except those used to test Gabor functions), we estimated the preferred features using natural images (12 × 12 pixels). Since natural images contain significant spatial correlations, it was necessary to modify the STC analysis in order to compute the preferred features. Details of this method can be found in Touryan et al. [27]. Briefly, we first corrected for the correlations in the stimuli by “whitening” each image in the ensemble:
where S is a vector representation of the stimulus (luminance in each pixel at each frame), U is a matrix containing the eigenvectors of the covariance matrix of S, and λ
1, …, n are the corresponding eigenvalues. As a result, Sw represents the stimulus in the whitened space. STC analysis for white-noise stimuli [25,26] was then applied to the ensemble Sw to identify significant eigenvectors (Vsig). The preferred features are then computed as
Since some eigenvalues (λ) for natural stimuli are very small, the “whitening” step will result in noise amplification. To solve this problem, a cutoff threshold was chosen such that whitening is performed only for eigenvectors above this cutoff (in this study, 50 out of 144 eigenvectors [27]).
STA analysis
This technique has been used previously to estimate the linear RFs of sensory neurons [23,36]. Briefly, to analyze the responses to white-noise stimuli, VSTA is calculated as the average of the spike-triggered stimulus ensemble. To analyze the responses to natural stimuli, the stimulus correlations are corrected for by normalizing VSTA by the covariance matrix of S (see STC analysis). To avoid noise amplification, this normalization was performed only above the cutoff (50/144), identical to that used for the modified STC analysis.
Although both STA and STC can be used to estimate preferred spatiotemporal features, in the present study we focused on the space domain in order to improve the signal-to-noise ratio of the estimate. In preliminary studies, we computed the spatiotemporal features of a subset of cells and found that the signals in the features are almost completely contained in the spatial feature at a delay of one frame (Figure S2). Thus in the present study the analysis was performed only at that delay. For each cell, we used the responses to 1–3 repeats of a natural-image ensemble to compute the preferred features.
Contrast-response function
The contrast of preferred feature k
φ (x, y) in stimulus S(x, y) is measured as the dot product of k
φ and S:
(see Figure 1C); k
φ satisfies
Since −1 ≤ S(x, y) ≤ 1, where 1 and −1 represent the highest and lowest luminance of the monitor, respectively, the above definition ensures that the feature contrast in each stimulus is bound between −1 and 1 (although, in practice, these limits were never reached with the stimulus ensembles used in this study). For each feature, the contrast-response function was measured from the neuronal responses to 1–3 repeats of an ensemble; for nearly all cells (38/40), these repeats were distinct from those used to estimate the preferred features, in order to avoid bias [29].
Analysis of feature detectability
For each visual feature and each stimulus ensemble, 50,000 trials were performed in silico for both the positive and negative values of the feature contrast. In each trial, a pair of neuronal responses (rp and ra), corresponding to “feature-present” and “feature-absent” stimuli respectively, were generated based on the probability distributions shown in Figure 6B (solid and dashed lines). The larger response of the pair was classified as the response to the “feature-present” stimulus (rp′); trials in which the responses were equal were excluded. The percentage of trials with correct classification (rp′ matches rp) was computed for both the positive and negative feature contrasts, and the results were averaged. The result shown in Figure 6C was qualitatively independent of the specific values of T
0 and T
1 over a wide range.
Supporting Information
Figure S1 Comparison between STA and Significant Eigenvector of STC for Simple Cells
Results from two cells are shown. Upper, STA. Middle, STC eigenvector. Lower, STC eigenvalues, as in Figure 1B.
(299 KB PDF).
Click here for additional data file.
Figure S2 Estimated Spatiotemporal Features of a Complex Cell
Note that the signals in the features are almost completely contained in the spatial features at a delay of one frame.
(278 KB PDF).
Click here for additional data file.
Protocol S1 Analysis of Simple-Cell Feature Sensitivity Using STC
(41 KB PDF).
Click here for additional data file.
Protocol S2 Global Contrast within RF of the Cell
(19 KB PDF).
Click here for additional data file.
We thank Michael Gastpar, Natalia Caporale, Avideh Zakhor, and Bobak Nazer for helpful discussions. This work was supported by a grant from the National Eye Institute (R01 EY12561).
Competing interests. The authors have declared that no competing interests exist.
Author contributions. GF, JT, and YD conceived and designed the experiments. GF, JT, and FH performed the experiments. GF and JT analyzed the data. GF, JT, and YD wrote the paper.
¤a Current address: Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, United States of America
¤b Current address: Department of Neurobiology, Yale School of Medicine, New Haven, Connecticut, United States of America
Citation: Felsen G, Touryan J, Han F, Dan Y (2005) Cortical sensitivity to visual features in natural scenes. PLoS Biol 3(10): e342.
Abbreviations
RFreceptive field
STAspike-triggered average
STCspike-triggered covariance
V1primary visual cortex
==== Refs
References
Attneave F Some informational aspects of visual perception Psychol Rev 1954 51 183 193
Field DJ Relations between the statistics of natural images and the response properties of cortical cells J Opt Soc Am A Opt Image Sci Vis 1987 4 2379 2394
Atick JJ Could information processing provide an ecological theory of sensory processing? Network 1992 3 213 251
Field DJ What is the goal of sensory coding? Neural Comput 1994 6 559 601
Thomson MG Beats, kurtosis and visual coding Network 2001 12 271 287 11563530
Geisler WS Perry JS Super BJ Gallogly DP Edge co-occurrence in natural images predicts contour grouping performance Vision Res 2001 41 711 724 11248261
Sigman M Cecchi GA Gilbert CD Magnasco MO On a common circle: Natural scenes and Gestalt rules Proc Natl Acad Sci U S A 2001 98 1935 1940 11172054
Barlow HB Rosenblith WA Possible principles underlying the transformation of sensory messages Sensory communication 1961 Cambridge, MA MIT Press 217 234
Simoncelli EP Olshausen BA Natural image statistics and neural representation Annu Rev Neurosci 2001 24 1193 1216 11520932
Zetzsche C Rohrbein F Nonlinear and extra-classical receptive field properties and the statistics of natural scenes Network 2001 12 331 350 11563533
Srinivasan MV Laughlin SB Dubs A Predictive coding: A fresh view of inhibition in the retina Proc R Soc Lond B Biol Sci 1982 216 427 459 6129637
Dong DW Atick JJ Temporal decorrelation: A theory of lagged and nonlagged responses in the lateral geniculate nucleus Network 1995 6 159 178
Dan Y Atick JJ Reid RC Efficient coding of natural scenes in the lateral geniculate nucleus: Experimental test of a computational theory J Neurosci 1996 16 3351 3362 8627371
Olshausen BA Field DJ Emergence of simple-cell receptive field properties by learning a sparse code for natural images Nature 1996 381 607 609 8637596
Bell AJ Sejnowski TJ The “independent components” of natural scenes are edge filters Vision Res 1997 37 3327 3338 9425547
van Hateren JH Ruderman DL Independent component analysis of natural image sequences yields spatio-temporal filters similar to simple cells in primary visual cortex Proc R Soc Lond B Biol Sci 1998 265 2315 2320
Hyvarinen A Hoyer PO A two-layer sparse coding model learns simple and complex cell receptive fields and topography from natural images Vision Res 2001 41 2413 2423 11459597
Kayser C Kording KP Konig P Learning the nonlinearity of neurons from natural visual stimuli Neural Comput 2003 15 1751 1759 14511511
Hubel DH Wiesel TN Receptive fields, binocular interaction and functional architecture in the cat's visual cortex J Physiol 1962 160 106 154 14449617
Movshon JA Thompson ID Tolhurst DJ Receptive field organization of complex cells in the cat's striate cortex J Physiol 1978 283 79 99 722592
Adelson EH Bergen JR Spatiotemporal energy models for the perception of motion J Opt Soc Am A Opt Image Sci Vis 1985 2 284 299
Fitzpatrick D Seeing beyond the receptive field in primary visual cortex Curr Opin Neurobiol 2000 10 438 443 10981611
Smyth D Willmore B Baker GE Thompson ID Tolhurst DJ The receptive-field organization of simple cells in primary visual cortex of ferrets under natural scene stimulation J Neurosci 2003 23 4746 4759 12805314
David SV Vinje WE Gallant JL Natural stimulus statistics alter the receptive field structure of v1 neurons J Neurosci 2004 24 6991 7006 15295035
Brenner N Bialek W de Ruyter van Steveninck R Adaptive rescaling maximizes information transmission Neuron 2000 26 695 702 10896164
Touryan J Lau B Dan Y Isolation of relevant visual features from random stimuli for cortical complex cells J Neurosci 2002 22 10811 10818 12486174
Touryan J Felsen G Dan Y Spatial structure of complex cell receptive fields measured with natural images Neuron 2005 45 781 791 15748852
Dean AF The relationship between response amplitude and contrast for cat striate cortical neurones J Physiol 1981 318 413 427 7320898
Chichilnisky EJ A simple white noise analysis of neuronal light responses Network 2001 12 199 213 11405422
Rust NC Schwartz O Movshon JA Simoncelli EP Spatiotemporal elements of macaque v1 receptive fields Neuron 2005 46 945 956 15953422
Maffei L Fiorentini A The unresponsive regions of visual cortical receptive fields Vision Res 1976 16 1131 1139 969225
Bonds AB Role of inhibition in the specification of orientation selectivity of cells in the cat striate cortex Vis Neurosci 1989 2 41 55 2487637
Maffei L Fiorentini A Bisti S Neural correlate of perceptual adaptation to gratings Science 1973 182 1036 1038 4748674
Schwartz O Simoncelli EP Natural signal statistics and sensory gain control Nat Neurosci 2001 4 819 825 11477428
Heeger DJ Normalization of cell responses in cat striate cortex Vis Neurosci 1992 9 181 197 1504027
Theunissen FE David SV Singh NC Hsu A Vinje WE Estimating spatio-temporal receptive fields of auditory and visual neurons from their responses to natural stimuli Network 2001 12 289 316 11563531
Chance FS Nelson SB Abbott LF Complex cells as cortically amplified simple cells Nat Neurosci 1999 2 277 282 10195222
Mechler F Ringach DL On the classification of simple and complex cells Vision Res 2002 42 1017 1033 11934453
Priebe NJ Mechler F Carandini M Ferster D The contribution of spike threshold to the dichotomy of cortical simple and complex cells Nat Neurosci 2004 7 1113 1122 15338009
Skottun BC De Valois RL Grosof DH Movshon JA Albrecht DG Classifying simple and complex cells on the basis of response modulation Vision Res 1991 31 1079 1086 1909826
Green DM Swets JA Signal detection theory and psychophysics 1966 New York Wiley
Vinje WE Gallant JL Sparse coding and decorrelation in primary visual cortex during natural vision Science 2000 287 1273 1276 10678835
Kayser C Salazar RF Konig P Responses to natural scenes in cat V1 J Neurophysiol 2003 90 1910 1920 12750423
Heeger DJ Simoncelli EP Movshon JA Computational models of cortical visual processing Proc Natl Acad Sci U S A 1996 93 623 627 8570605
Yu Y Lee TS Adaptive contrast gain control and information maximization Neurocomputing 2005 65–66 111 116
Vinje WE Gallant JL Natural stimulation of the nonclassical receptive field increases information transmission efficiency in V1 J Neurosci 2002 22 2904 2915 11923455
Burr D Morrone C Maffei L Intra-cortical inhibition prevents simple cells from responding to textured visual patterns Exp Brain Res 1981 43 455 458 7262240
Muller JR Metha AB Krauskopf J Lennie P Local signals from beyond the receptive fields of striate cortical neurons J Neurophysiol 2003 90 822 831 12724358
Morrone MC Burr DC Feature detection in human vision: A phase-dependent energy model Proc R Soc Lond B Biol Sci 1988 235 221 245 2907382
Mechler F Reich DS Victor JD Detection and discrimination of relative spatial phase by V1 neurons J Neurosci 2002 22 6129 6157 12122074
Oppenheim AV Lim JS The importance of phase in signals Proc IEEE 1981 69 529 541
Piotrowski LN Campbell FW A demonstration of the visual importance and flexibility of spatial-frequency amplitude and phase Perception 1982 11 337 346 7167342
Wang Z Simoncelli EP Thrun S Saul L Scholkopf B Local phase coherence and the perception of blur Advances in neural information processing systems 2003 Cambridge, MA MIT Press
|
16171408
|
PMC1233414
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e342
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030342
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030348SynopsisBiotechnologyCell BiologyMolecular Biology/Structural BiologyNeurosciencePhysiologyDermatologyBiochemistryHomo (Human)Mus (Mouse)Fibroblast Culture Cells Keep Track of Circadian Rhythms Synopsis10 2005 27 9 2005 27 9 2005 3 10 e348Copyright: © 2005 Public Library of Science.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 Period Length of Fibroblast Circadian Gene Expression Varies Widely among Human Individuals
==== Body
Embedded in the brain's hypothalamus on top of the intersection of the two optic nerves lies a group of some 10,000 cells known as the suprachiasmatic nucleus. This neural region determines the biological equivalent of Greenwich Mean Time: the circadian cycle. Circadian rhythms influence phenomena as diverse as leaf position in plants, fatigue patterns in mice, and late-night hunger in human adolescents. By measuring exactly how long an organism takes to complete one cycle, scientists hope to gain insight into the mechanisms underlying these broad effects. “Circadian” translates from Latin to “approximately a day,” but scientists search for more precise answers about how genes and environment account for fluctuations in circadian cycle length.
Theoretically, scientists could measure the circadian signal from its source in the suprachiasmatic nucleus—but no method exists to do this in a living subject. Instead, scientists rely on prolonged observation of live mice or humans. They document when mice use their mouse wheel, for example, and when they sleep. But to screen for and identify circadian rhythm variations in humans, the required period of lengthy observation is prohibitively costly and labor intensive.
Studying circadian rhythms in humans isn't as straightforward as it is with lab rodents. As a proxy for humans, researchers compared circadian gene expression in primary fi broblasts from different human individuals and found unexpectedly large differences between the circadian clocks of different human subjects
Circumventing these technical limitations, Ueli Schibler and colleagues have recently designed a method to measure circadian cycles in mammalian cells cultured from tissues other than the suprachiasmatic nucleus. The researchers infected human and mouse fibroblast tissue cultures with a virus engineered to report when a certain host circadian rhythm gene was expressed. They found that their data jibed with the previously accepted length of the human circadian cycle: 24.5 hours. Because of the sensitivity of their method, Schibler and colleagues also confirmed that, for both humans and mice, circadian rhythms vary substantially between individuals. This suggests that the genetics of the circadian clock likewise varies between individuals.
The so-called reporter gene, lucerifase, produces a protein that illuminates the cell. Steve Brown, a postdoc in Schibler's laboratory, created a virus system that specifically inserts lucerifase into the host genome near a gene important for establishing the circadian rhythm. With this system, the same gene that promotes expression of the circadian rhythm gene also promotes expression of lucerifase—consequently, the fibroblast cultures light up in time with the circadian clock.
Schibler and colleagues obtained human fibroblast skin cells from the abdomen, buttocks, male foreskin, and other areas to measure the circadian rhythm indirectly. Unlike an individual's true rhythm, a fibroblast cell culture's rhythm does not vary with changes in light exposure or sleep habits. The researchers point out that their method can expose differences in circadian rhythms; it does not, however, directly measure the signal from the suprachiasmatic nucleus that coordinates each individual cell's clock.
The authors also found that circadian rhythm variations between cultures from the same individuals differed far less than differences between individuals. This discovery indicates that fibroblast cell cultures are reliable tools for approximating an individual's genetically determined circadian rhythm. The scientists also found that, for mice, circadian rhythm time obtained from prolonged observation and that obtained from the new viral method were showing the same tendency, although the differences observed in behavior were exacerbated in molecular fibroblast rhythms. This result suggests that the fibroblast circadian clocks might give relevant information about the brain's circadian clock in humans as well as in mice.
In the future, scientists may use the new method to screen large populations for genetically linked sleep disturbances such as advanced and delayed sleep phase syndromes. They may also use this test to home in on the genetic mechanism responsible for such conditions. Outside the realm of medicine, future genetic studies of circadian rhythm may exploit the method developed by Schibler and colleagues to explore questions about the exact relationship between the suprachiasmatic nucleus and the circadian rhythms of cultured fibroblast cells. Just how does this brain structure coordinate genetic imperatives with environmental input including light fluctuations? Finally, frequent fliers, bleary-eyed in foreign time zones, may get an answer to why waking up is so hard to do. —Jessica Tanenbaum
|
0
|
PMC1233415
|
CC BY
|
2021-01-05 08:21:28
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e348
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030348
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030359SynopsisGenetics/Genomics/Gene TherapyMicrobiologyBiochemistryEubacteriaMapping Core Communication Networks in Bacteria Synopsis10 2005 27 9 2005 27 9 2005 3 10 e359Copyright: © 2005 Public Library of Science.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.
Two-Component Signal Transduction Pathways Regulating Growth and Cell Cycle Progression in a Bacterium: A System-Level Analysis
==== Body
Single-celled bacteria may appear simple by some standards, but these tiny cells employ sophisticated systems for processing stimuli. One especially important class of signaling molecules that help bacteria coordinate the activities of daily life is called the two-component signal transduction system. This system—comprised of enzymes called histidine kinases and their target molecules, the response regulators—allows bacteria to sense and respond to their surroundings by transforming various environmental cues, such as sugars, peptides, and antibiotics, into physiological responses. These environmental signals trigger a chemical reaction in histidine kinases called autophosphorylation, in which the kinase transfers a phosphoryl group from a molecule of ATP (which powers many cellular processes) to one of its own amino acids. The enzyme then transfers the phosphoryl group to its target response regulator, producing changes in gene expression, motility, protein breakdown, and various other cellular processes.
Though it's possible to identify histidine kinases and response regulators in bacteria by analyzing their genome sequences, it's far more difficult to determine how the two components interact: do they form monogamous pairs or behave more promiscuously? To characterize the range of possible interactions and the intracellular changes they bring about, Michael Laub and colleagues developed a novel method of mapping the connections between histidine kinases and response regulators in the freshwater bacterium Caulobacter crescentus. By combining genetic and biochemical analyses on a system-wide scale, the authors rapidly identified two-component signaling pathways in C. crescentus, including pathways required for core cell processes.
In the two-component signaling pathway, a stimulus causes a histidine kinase to autophosphorylate on a conserved histidine (H) residue. The phosphoryl group (P) is then transferred to a conserved aspartate (D) residue of a cognate response regulator; when the response regular phosphorylates, it triggers a physiological response, such as gene transcription
Laub and colleagues first identified 106 two-component signal transduction genes (62 histidine kinases and 44 response regulators) from the genome sequence. Then they created mutant strains of bacteria that each had one of the 106 genes deleted (called deletion strains) to learn how the genes function. The phenotypes, or physical characteristics, of the mutant strains allowed the scientists to identify 39 genes required for cell cycle progression, growth, and morphogenesis, including nine genes essential to survival. To address the promiscuity question and figure out the likely phosphotransfer pairings among the components, the authors developed a biochemical method, called phosphotransfer profiling, that quickly identifies a histidine kinase target by tracking the transfer of radioactive (called radio-labeled) phosphates from the kinase to the target.
The authors validated their in vitro technique on two-component proteins from E. coli (a system in which many of the living bacteria's kinase-target pairings are known) by showing that the kinases preferentially phosphorylated their known targets in the test tube as well, forming promiscuous unions only after prolonged periods. Confident that their method would also work for other bacteria, the researchers applied it to C. crescentus and determined the likely pairings for previously identified histidine kinases. Since the authors again observed a high preference for known targets, they were confident that these pairings were real, and the method could be used to identify targets of other, uncharacterized kinases. Based on the deletion analysis results, Laub and colleagues focused on a histidine kinase that appears essential for growth or survival and identified a single target response regulator. Eliminating the expression of each gene produced nearly identical phenotypes, revealing the pair's role in a signaling pathway that maintains the integrity of the bacteria's cellular envelope. Because two-component systems aren't found in humans, this cellular-envelope pathway may prove an effective antibiotic target against pathogenic bacteria, based on the effectiveness of other antibacterial therapies aimed at the cell membrane.
The library of deletion strains the authors developed will be a valuable resource, not only for identifying other two-component pathways in C. crescentus but also for studying aspects of the cell cycle and development, thanks to the bacterium's unusual life cycle—it divides asymmetrically, producing a motile daughter cell and a stationary one that can then differentiate into the motile version. And since the techniques presented here should work in any organism with two-component signaling (most bacteria, fungi, and many plants), researchers can apply them to the tall task of decoding the labyrinthine communication systems that sustain cellular life. —Liza Gross
|
0
|
PMC1233416
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e359
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030359
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030360SynopsisNeuroscienceCatComplex Cells in the Brain's Vision Center Tune in to Natural Scenes Synopsis10 2005 27 9 2005 27 9 2005 3 10 e360Copyright: © 2005 Public Library of Science.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.
Cortical Sensitivity to Visual Features in Natural Scenes
==== Body
An Amur tiger roaming the snow-covered forests of the Russian Far East sees life differently than an arboreal primate raised in the thick canopy of a tropical jungle. Adaptations in the structure of the eye and the visual centers of the brain facilitate these different worldviews, fine-tuning each animal's vision to the light levels and visual properties of its environment.
The notion that brain circuits are adapted to represent natural stimuli is known as the efficient coding hypothesis. In this framework, the visual system responds most effectively to the features found in the natural environment, like the specific arrangement of trees, another animal's face, or the contrast of land abutting a river. Natural images possess different statistical features than random noise (such as static on a TV monitor), so responses to these two classes of stimuli could be distinct. According to the prevailing hypothesis, the retina and the lateral geniculate nucleus, the brain area that relays neural signals to the cortex, are tuned to the power spectrum of light signals found in natural scenes. In other words, these neurons are sensitive to the dominant energy at particular spatial or temporal frequencies in natural scenes and less sensitive to random noise, which has an overall flat power spectrum.
In a new study, Gidon Felsen, Yang Dan, and colleagues investigate how neurons in the cat primary visual cortex (V1) respond to images with natural statistics, and discover something truly novel about a V1 neuron's sensitivity to features in natural scenes: a specific class of V1 neurons, called complex cells, are preferentially tuned to the phase regularities of light signals in natural scenes, not to the power spectrum. (Phase relates to how the signals in different spatial frequencies align, which gives rise to edges in the image.)
By recording cortical responses to several classes of natural and synthetic images, researchers showed that complex cells are tuned to the phase structure of natural images to represent image features, such as the edges highlighted on the clock tower, efficiently
The V1 contains two types of neurons, known as simple and complex cells, that were originally distinguished based on how they respond to light, determined by shining a flashlight on a wall and mapping the space that triggered neuronal activation. This activation space defines the cell's receptive field. The receptive field properties of simple and complex cells vary significantly: simple cells behave linearly (two spots of light in the receptive field doubles the response), and complex cells behave nonlinearly.
To characterize the feature sensitivity of these neurons, the authors measured their response to different classes of visual stimuli, including natural, random, and synthesized images; the synthesized images helped them distinguish between the power spectrum and phase effects. Feature sensitivity depends on a neuron's preferred features, which the authors estimated from a neuron's response to a set of natural images, including a man's face, a building, a lion, and a hand. The authors created a set of random images with global and feature contrasts that matched the natural images, based on the preferred features, and recorded neuronal response to both the natural and random image sets. Overall, the majority of complex cells showed higher feature sensitivity for the natural stimuli, indicated by the contrast-response function, which plots neuronal response against the contrast of the feature. (A steep contrast-response function shows high neuronal sensitivity to a preferred feature, while a flat function indicates insensitivity.) Simple cells showed no difference in their response to natural and random stimuli.
Since the contrasts of natural and random image sets were matched, the complex cells' sensitivity to natural images could not be explained by differences in overall contrast. More likely, the authors reasoned, the cells were responding to the power or phase spectra of the light. Felsen et al. manipulated each property in synthesized image sets to distinguish their effects on neuronal response. In one image set, each image had a natural power and random phase spectrum, while a second image set had the reverse. Comparing the feature sensitivity of complex cells to each of these image sets with a random image set, the contrast-response function, and thus feature sensitivity, was highest for the synthesized natural phase image set.
By experimentally linking visual statistics with neuronal responses, this study not only reveals a novel coding response property of complex cells but also provides evidence for the theory of efficient coding. The finding that complex cells selectively respond to properties of natural stimuli that simple cells don't shows how brain circuits divide tasks to make the most of available resources. Researchers can now investigate how the structure of complex cells engenders their heightened sensitivity. —Liza Gross
|
0
|
PMC1233417
|
CC BY
|
2021-01-05 08:21:28
|
no
|
PLoS Biol. 2005 Oct 27; 3(10):e360
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030360
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1612235110.1371/journal.pbio.0030312Research ArticleCell BiologyDevelopmentSystems BiologyParasitologyNematodesCaenorhabditisFunctional Genomic Analysis of C. elegans Molting Genomics of C. elegans MoltingFrand Alison R
1
Russel Sascha
1
Ruvkun Gary [email protected]
1
1Department of Molecular Biology, Massachusetts General Hospital, Boston, Massachusetts, United States of America, and Genetics Department, Harvard Medical School, Boston, Massachusetts, United States of AmericaPlasterk Ronald Academic EditorUtrecht UniversityNetherlands10 2005 30 8 2005 30 8 2005 3 10 e3123 3 2005 7 7 2005 Copyright: © 2005 Frand 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 Worm Sheds Its Genetic Secrets
Although the molting cycle is a hallmark of insects and nematodes, neither the endocrine control of molting via size, stage, and nutritional inputs nor the enzymatic mechanism for synthesis and release of the exoskeleton is well understood. Here, we identify endocrine and enzymatic regulators of molting in C. elegans through a genome-wide RNA-interference screen. Products of the 159 genes discovered include annotated transcription factors, secreted peptides, transmembrane proteins, and extracellular matrix enzymes essential for molting. Fusions between several genes and green fluorescent protein show a pulse of expression before each molt in epithelial cells that synthesize the exoskeleton, indicating that the corresponding proteins are made in the correct time and place to regulate molting. We show further that inactivation of particular genes abrogates expression of the green fluorescent protein reporter genes, revealing regulatory networks that might couple the expression of genes essential for molting to endocrine cues. Many molting genes are conserved in parasitic nematodes responsible for human disease, and thus represent attractive targets for pesticide and pharmaceutical development.
The authors use a genome-wide RNA-interference screen to identify and characterize genes involved in C. elegans molting. They investigate regulatory networks involved in molting, lending important new insights into this complex process.
==== Body
Introduction
Ecdysozoan animals, including nematodes and arthropods [1], develop through periodic larval stage molts when the exoskeleton is shed and synthesized anew. Although molting is the hallmark of the most abundant and diverse group of animals on the planet, including a wide variety of human pests and pathogens, the endocrine circuits that regulate molting in response to environmental and physiologic cues are not well understood. Moreover, little is known about the molecular mechanisms for release and de novo production of the exoskeleton.
Endocrine and neuroendocrine pathways regulate molting in arthropods, and likely operate in nematodes as well. In insects, pulses of the steroid hormone ecdysone trigger molting and metamorphosis [2,3]. The neuropeptide prothoracicotropic hormone stimulates synthesis of ecdysone in the prothoracic glands [3]. At the end of each larval stage, the neuropeptide eclosion hormone, combined with a decline in the titer of ecdysone, prompts release of the peptide ecdysis-triggering hormone from glands lining the trachea [4–7]. Ecdysis-triggering hormone then promotes behaviors essential for escaping the old exoskeleton [8,9], and also stimulates neurons to secrete more eclosion hormone, creating a positive feedback loop that culminates in a hormonal surge decisive for ecdysis [10]. Environmental cues, including photoperiod, temperature, and humidity, as well as physiologic factors, including size, stage, and the nutritional status of the organism, modulate secretion of prothoracicotropic hormone in various arthropods, suggesting extensive sensory input to the neuroendocrine secretions that govern molting [3]. However, little is known about the circuits that initiate, terminate, or set the pace of the molting cycle in any Ecdysozoan.
Although an endocrine trigger for nematode molting has yet to be identified, several lines of evidence implicate steroid hormones in Caenorhabditis elegans molting. Molting of C. elegans requires cholesterol, the biosynthetic precursor of all steroid hormones, as well as the low-density lipoprotein (LDL) receptor-like protein LRP-1, which is thought to endocytose sterols from the growth medium [11]. A sterol-modifying enzyme synthesized in the intestine, LET-767, is also essential for molting, consistent with the production or modification of a hormone derived from steroids [12]. The best evidence of a hormonal cue for molting of C. elegans is the requirement for two nuclear hormone receptors (NHRs), NHR-23 and NHR-25, orthologous, respectively, to the ecdysone-responsive gene products DHR3 and Ftz-F1 of Drosophila melanogaster [13–16]. Ecdysone itself, however, is unlikely to serve as a molting hormone in nematodes because ecdysteroids have not been detected in any free-living nematode [17], and because orthologs of the ecdysone receptor components EcR and USP (Ultraspiracle) have not been identified in the complete genome of C. elegans [18].
Molting of C. elegans involves the synthesis and secretion of a new exoskeleton underneath the old one, separation of the old exoskeleton from the epidermis (apolysis), and escape from the old exoskeleton (ecdysis) [19]. At the end of each stage, larvae become inactive for a brief period of time known as lethargus that coincides with separation of the old exoskeleton from the epidermis. Next, particular behaviors promote ecdysis; larvae flip on their long axis to loosen the body cuticle, expel the anterior half of the pharyngeal cuticle, and ultimately escape the old exoskeleton via a forward thrust [19].
The exoskeleton of nematodes, called the cuticle, is a collagenous extracellular matrix secreted by underlying epithelial cells, known as the hypodermis and seam cells, and also by specialized interfacial cells that line openings of the body, including the buccal cavity, pharynx, vulva, rectum, and sensilia [20]. Lipids and glycolipids comprise the outermost layer of the cuticle, whereas glycoproteins, thought to be secreted by gland cells, form the surface coat [21]. Elasticity of the cuticle permits growth during each larval stage, but particular structures, such as the buccal cavity, grow saltationally at molts [22]. The distinction between collagen in the nematode exoskeleton and chitin in the insect exoskeleton suggests that the enzymatic cascades that mediate release of the exoskeleton in nematodes may be distinct from those that release the exoskeleton in arthropods. Although two collagenases essential for molting have been identified in C. elegans [23,24], the full ensemble of signaling proteins and extracellular matrix enzymes required to remodel the exoskeleton has yet to be illuminated.
Human diseases caused by parasitic nematodes affect tropical regions of Africa, Asia, and South America. The World Health Organization estimates that 120 million people endure lymphatic filariasis (elephantiasis), due to infection by the filarial nematodes Wuchereria bancrofti or Brugia malayi, and that 18 million people endure onchocerciasis (African river blindness), due to infection by Onchocerca volvulus [25]. Ascaris, hookworms, and whipworms are also important pathogens, infecting approximately 1 billion people. Parasitic nematodes further damage livestock and lay waste to $80 billion of crop plants annually. One promising approach to the discovery of new targets for anti-nematode drugs, vaccines, and pesticides is the identification of nematode-specific genes essential for the viability of larvae. In the screen described here, a large number of nematode-specific genes essential for molting were identified, and some encode attractive drug targets.
Here, we identify endocrine and enzymatic regulators of molting in C. elegans through a genome-wide RNA-interference (RNAi) screen, providing a broad view of functions essential for molting in a model Ecdysozoan. We further develop models for the genetic regulation of molting based on the location, timing, and order of expression of particular molting genes.
Results
To identify a full set of endocrine and enzymatic regulators of molting in C. elegans, we screened a combined library of 18,578 bacterial clones that each express a double-stranded RNA designed to silence one of the 19,427 predicted worm genes via RNAi [26–28]. About 25 L1-stage larvae were fed each clone and later examined for molting defects, indicated by the adherence of cuticle from the pre-molt larval stage to the body of the worm (the Mlt phenotype; Figure 1). Gene inactivations observed to prevent molting in the primary library screen were tested again by feeding the bacterial clones to approximately 50 wild-type (N2) and 50 rrf-3(pk1426) mutant larvae, a genetic background where RNAi is more effective [29].
Figure 1 Molting Defects Caused by RNAi
N2 larvae were fed bacteria expressing dsRNA corresponding to the indicated genes, or control bacteria not expressing dsRNA of a worm gene (A). Panels A–D show the anterior, whereas (E) shows the mid-body of a larva. Black arrowheads mark unshed cuticle. White arrowhead (C) indicates the buccal capsule. Nomarski optics.
Inactivation of 159 genes (Tables 1 and S1–S4) interfered with molting. Tables 1 and S1 show genes whose inactivation produced molting defects in 10% to 100% of wild-type (N2) or rrf-3(pk1426) mutant larvae. Eighty-seven other genes were assigned a lower priority based on gene annotation (Tables S2 and S3) or the low penetrance of molting defects observed after RNAi (Table S4). The blind identification of nine genes previously described to cause an arrest at a molt, including lrp-1, nhr-23, nhr-25, nas-37, nas-36, skp-1, rme-8, acn-1, and bli-4 [11,13–16,23,24,30–33], verified the efficacy of this RNAi-based strategy for isolating bona fide molting genes. An additional 28 of the 159 gene inactivations were independently described as causing an arrest at a molt in broad screens that identified many loss-of-function phenotypes using RNAi [26,27,34]. Further, the observation of molting defects in larvae with mutations in qua-1 (unpublished data), lrp-1 [11], and nas-37 [23] verified that RNAi faithfully recapitulates loss-of-function phenotypes in the molting pathway. The names mlt-8, mlt-9, and mlt-11 were assigned, respectively, to W08F4.6, F09B12.1, and W01F3.3 after expression data verified a primary function in molting. The genes fbn-1, noah-1, and noah-2 were assigned names based on homology to genes of mammals or insects (Table S1).
Table 1 Selected Genes Whose Inactivation Disrupts Molting
Figure 1 shows examples of molting-defective larvae produced by RNAi. Most often, larvae were observed incarcerated in sheaths of old cuticle extending from the anterior end of the worm, as shown for a mlt-11(RNAi) larva (Figure 1B). The nature of molting defects caused by particular gene inactivations suggested that the corresponding proteins function in a specific anatomical place or stage of ecdysis. For example, L4-stage larvae fail to shed cuticle from the pharynx after RNAi of xrn-2 (unpublished data). A similar type of molting defect has been observed in animals lacking the DNA binding protein PEB-1 [35]. Many mlt-9(RNAi) larvae fail to shed cuticle lining the buccal cavity, causing the lips to evert (Figure 1C). Unshed cuticle often forms coronal constrictions on nas-37(RNAi) larvae (Figure 1E), possibly when animals flip on their long axis during molting. Inactivation of the collagenase gene nas-37 also prevents the breakdown of old cuticle at the anterior tip of the worm, thereby blocking escape from the old exoskeleton (unpublished data) [23,24]. Although Mlt larvae typically arrest development, some gradually escape from the old cuticle, only to fail again at the next molt, a phenomenon observed often after RNAi of qua-1 (unpublished data).
The majority of genes we identified are likely to act at all four molts, because their inactivation prevents molting from several larval stages. Moreover, although feeding L1-stage larvae dsRNA for particular genes, such as mlt-8 or acn-1, prevents development beyond the L3 stage (Table S5), feeding the same dsRNAs to older larvae also disrupts the final molt (unpublished data). The majority of gene inactivations also disrupt molting from the dauer stage, an alternative L3 stage that is adapted for survival in unfavorable conditions and resembles the infective form of parasitic nematodes (Table S6).
The majority of genes we identified are conserved in parasitic nematodes responsible for human, animal, and plant diseases (Table S7). Many of the genes, including mlt-8 and mlt-9, are conserved only in nematodes; similar proteins are readily identified among the predicted products of cDNAs or genomic sequences from parasitic and free-living nematodes (Table S7), but not in the translated genomes of D. melanogaster or Homo sapiens (Table S1). In contrast, the genes noah-1 and noah-2, which specify putative extracellular matrix components, are conserved in insects and nematodes, but not in humans, and thus show the phylogenetic conservation signature expected for molting genes common to Ecdysozoans.
Predicted Functions of Genes Uncovered in the Molting Screen
In this section, we discuss how the annotations of particular genes uncovered by RNAi implicate the corresponding proteins in basic aspects of the molting cycle. Based on experimental evidence of a steroidal pathway, as well as the evolutionary relationship between arthropods and nematodes [1], we expect that endocrine cues periodically initiate molting in C. elegans, stimulating the synthesis of a new cuticle and release of the old one. We expected to isolate many genes essential for apolysis or ecdysis because we screened for larvae arrested at the final stage of the molt. However, we also anticipated the identification of genes required for the production of, or response to, hormonal cues for molting, because the loss of either nhr-23 in C. elegans or EcR in D. melanogaster can result in a terminal failure to ecdyse [14,36]. Also, a breakdown in the coordination of signaling events associated with molting might trigger an aberrant ecdysis and thereby cause arrest at that stage.
Regulation of Gene Expression
The identification of several transcription factors suggests that molting of C. elegans requires extensive changes in gene expression, similar to how transcriptional cascades promote molting and metamorphosis of insects [2]. Particular transcription factors likely alter gene expression in epithelial cells, possibly in response to endocrine cues. Annotated DNA binding proteins and transcription factors required for molting include three zinc-finger proteins, specified by F10C1.5, F25H8.6, and lir-1 [37], that resemble, respectively, Drosophila Doublesex, BED subfamily members, and the C. elegans transcription factor LIN-26, as well as two NHRs, NHR-23 and NHR-25, that were previously implicated in molting [13,15]. NHR-23 and NHR-25 are the best candidates for transducing hormonal signals, because the NHRs are expressed in epithelial cells and conserved in insects [13,15,16,30]. In theory, NHRs required for molting might regulate the expression of zinc-finger transcription factors identified in this screen, just as particular NHRs activate zinc-finger proteins in the transcriptional cascades coupled to insect metamorphosis [2].
The xrn-2 gene encodes a 5′-3′exoribonuclease that is conserved from yeast to humans [38] and is essential for molting in C. elegans. The homologous enzyme, Rat1p, is required for degradation of nuclear pre-mRNAs as well as the 5′ processing of ribosomal and small nuclear RNAs in Saccharomyces cerevisiae [39–41]. Consistent with a role for C. elegans XRN-2 in gene regulation, the degradation pathway mediated by Rat1p in yeast is known to compete with productive mRNA splicing [41].
Intercellular Signaling
Probable signaling components were identified in the molting screen, consistent with expectations of an endocrine cue for molting, coordination of the process in different cell types, and physiologic feedback on the status of the molt to endocrine regulators. Putative signaling peptides include MLT-8, PAN-1, and QUA-1. Features of peptide hormones present in the novel protein MLT-8 include an N-terminal secretory signal sequence, two pairs of basic amino acids suitable for proteolytic processing, and three putative N-linked glycosylation sites. The predicted MLT-8 protein also lacks motifs characteristic of association with membranes or the extracellular matrix. Thus, we expect MLT-8 to be secreted from cells where it is synthesized and to serve as a signaling molecule. Cells might also secrete PAN-1, based on predictions of an N-terminal secretory signal sequence and putative glycosylation sites.
The qua-1 (quahog) gene specifies a protein with a Hint domain at the C-terminus [42], a hallmark of the hedgehog family of membrane-associated intercellular signaling proteins that suggests autocatalytic cleavage [43]. The predicted QUA-1 protein possesses a secretory signal sequence, but lacks obvious sequence homology to the N-terminal domain of hedgehog, which is active in signaling [43]. The genes ptr-4 and ptr-23 encode proteins similar to the transmembrane transporter Dispatched [43] that are also essential for molting (Table 1) and might export QUA-1 from cells where the protein is synthesized.
The acn-1 gene encodes a protein whose central region is 28% identical to human angiotensin converting enzyme (ACE)[32], the peptide protease that cleaves angiotensin I to angiotensin II. One model for the function of ACN-1 in molting is that ACN-1 regulates the production of a peptide molting hormone. However, ACN-1 is unlikely to directly catalyze proteolysis, because the active-site residues that coordinate zinc in human ACE are not conserved [32]. Nevertheless, ACN-1 might bind particular peptides, and thereby influence their maturation or secretion.
Protein Synthesis
The isolation of 25 genes encoding ribosomal proteins or tRNA synthetases (Table S3) confirmed that molting requires a burst of biosynthetic activity, presumably to make components for the new cuticle [19].
Secretion of the New Cuticle
Eighteen components of the general secretion machinery isolated in our screen (Table S2) are likely essential for synthesis of the new cuticle, including the vesicle coat proteins Sec-23p and B-cop, the small GTPase Sar-1p, and the vesicle fusion factor NSF [44, 45]. Consistent with a defect in synthesis of the new cuticle, the bodies of larvae undergoing RNAi of secretory genes often disintegrate at the molt, whereas the bodies of other Mlt larvae remain intact (unpublished data). Alternatively, defects in the secretory or endocytic trafficking of particular proteases or transmembrane proteins might account for molting defects caused by the loss of particular secretory pathway genes, similar to how the loss of the cytoplasmic adaptor protein DAB-1 interferes with molting, likely by disrupting intracellular transport of LRP-1 [46].
Remodeling of the Exoskeleton
Many genes identified here as essential for molting encode proteins predicted to directly regulate the production or release of the collagenous cuticle. Predicted components of the cuticle include FBN-1, a protein that is 30% identical to fibrillin, the microfibril protein defective in Marfan syndrome, a common disorder of connective tissue in humans [47]. In addition, the genes noah-1 (nompA-homolog) and noah-2 encode proteins homologous to NompA, a component of specialized extracellular matrices in the fly [48]. Identification of fbn-1, noah-1, and noah-2 as essential for molting suggests that incorporation of the corresponding proteins into macromolecular structures within the new cuticle might be critical for release of the cuticle at the next molt. We further identified three peroxidases likely to modify cuticle components, one of which, BLI-3, is thought to crosslink cuticle collagens [49]. Enzymatic modifications that occur after secretion of the cuticle might therefore be essential for the structural integrity of the new cuticle or shedding of the cuticle at the subsequent molt.
NAS-37 and NAS-36, two tolloid family metalloproteases independently described as essential for ecdysis [23,24], likely degrade the cuticle of the pre-molt larval stage, or regulate the maturation of other zymogens, as in the blood clotting protease cascade. NAS-37 and NAS-36 might also regulate the assembly of new cuticle [23] by processing the precursors of particular extracellular matrix proteins, just as tolloid family metalloproteases in vertebrates regulate extracellular matrix formation, in part, by cleaving the C-propeptides of procollagens [50,51]. The products of mlt-11 and bli-5 likely serve as extracellular protease inhibitors, because they contain multiple domains similar to bovine pancreatic trypsin inhibitor, and because a related protein from D. melanogaster, Papilin, localizes to the extracellular matrix in vivo and inhibits ADAMTS metalloproteases in vitro [52]. Identification of these probable anti-proteases suggests that proteolysis of the cuticle during molting is highly regulated, either spatially or temporally. In the absence of MLT-11 and BLI-5, extracellular proteases might be overly active in particular spatial domains, or at inappropriate times, a view consistent with the observation of blisters in the adult cuticle of bli-5(RNAi) animals (unpublished data).
The molting RNAi screen further identified transmembrane proteins expected to localize to the surface of epithelial cells and transduce signals that coordinate remodeling of the exoskeleton. The presence of MAM domains in the products of mlt-9 and ZC13.3 suggests a role in signaling, because the MAM domain is found in numerous transmembrane proteins of the cell-adhesion superfamily, including the receptor-like protein tyrosine phosphatase mμ [53]. The lrp-1 gene, one of the first identified as essential for molting [11], and also a prominent hit in the genomic screen, encodes an LDL receptor-like protein whose intracellular domain has been proposed to function as a signaling molecule after proteolytic cleavage and release from the apical membrane of epithelial cells [54].
Taken together, the loss-of-function phenotypes and annotations of these genes suggest that the corresponding proteins regulate the formation, condensation, or degradation of the collagenous cuticle during molting. Precise transcriptional regulation of these genes as well as post-translational regulation of the corresponding gene products is likely required for the orderly synthesis and breakdown of cuticle during each molt.
In addition to the genes uncovered by RNAi, we identified the mlt-10 gene as a hypermorphic allele in a forward genetic screen for molting mutants to be described elsewhere. The mlt-10 gene corresponds to C09E8.3 and represents the founding member of a large family of nematode-specific genes encoding putative membrane proteins or components of the cuticle (unpublished data).
Remodeling of Attachments between the Muscle, Hypodermis, and Exoskeleton
C. elegans move by the transmission of force from the contractile apparatus of muscle to the exoskeleton via a series of lateral attachments comprised of the dense bodies and M-lines of muscle, the basement membrane situated between the muscle and the hypodermis, and hemidesmosome-like structures of the hypodermis. Consistent with the view that hemidesmosomes are remodeled at the molt, our screen identified myotactin and MUP-4 (Table S1), components of the hemidesmosomes that link, respectively, the basal membrane of the hypodermis to the basement membrane of the muscle [55], and the apical membrane of the hypodermis to the inner layer of the cuticle [56]. Surprisingly, we also isolated the muscle protein tropomyosin [57] and the basement membrane protein UNC-52 [58], suggesting that muscle attachment points might also be remodeled at the molt. Intercellular signaling involving myotactin might guide remodeling of the connections between body wall muscle and the hypodermis in much the same way myotactin maintains the association between muscle and hypodermal fibrous organelles during embryogenesis [55,59].
Temporal and Spatial Expression Patterns of Molting Genes
We determined the spatial and temporal expression pattern of particular molting genes. We expected some of these genes to act in endocrine cells that trigger molting and some to act in epithelial cells that are remodeled during molting. Further, we anticipated much dynamic regulation during the molting cycle. Because the period between molts is short, only 8–10 h at 25 °C, we fused a PEST (Pro-Glu-Ser-Thr) signal for rapid protein degradation [60] to the C-terminus of green fluorescent protein (GFP), generating a fusion protein with a fluorescent half-life of less than 1 h in vivo. The gfp-pest gene was placed under the control of promoters from six genes, nas-37, mlt-11, mlt-9, acn-1, mlt-8, and mlt-10. Fusions with the conventional gfp gene revealed the cellular patterns, but not the detailed temporal dynamics, of expression from the promoters of qua-1 and xrn-2. The genes selected for analysis encode proteins that represent the major functional categories identified in this screen; namely, regulators of gene expression, including the exoribonuclease XRN-2; putative signaling pathway components, including the hedgehog-like QUA-1, the novel secreted peptide MLT-8, and the ACE homolog ACN-1; and annotated transmembrane or cuticle proteins, including the MAM domain protein MLT-9, the protease inhibitor MLT-11, the collagenase NAS-37, and the novel protein MLT-10. Further, each gene selected for expression studies produced penetrant molting defects after RNAi, or, in the case of mlt-10, after mutation.
Fluorescence from all eight of the gfp fusion genes was observed in epithelial cells that synthesize cuticle (Figures 2 and S1). The qua-1, nas-37, mlt-9, mlt-11, acn-1, mlt-8, and mlt-10 fusion genes were each expressed in the hypodermis, including the major body syncytium, hyp7, and hypodermal cells in the head and tail (Figure 2A–C and 2E–F) (unpublished data). The nas-37, mlt-9, acn-1, and mlt-11 fusion genes were also expressed in the lateral seam cells (Figure 2B–C and 2E–F), which are essential for molting and morphogenesis of the cuticle [61]. Interestingly, fluorescence from nas-37p::gfp-pest in seam cells was observed only before the L4 stage-to-adult molt, when the cells terminally differentiate and fuse, whereas fluorescence from mlt-9p::gfp-pest and mlt-11p::gfp-pest in seam cells was observed, predominantly, before larval-to-larval molts, when the cells divide. The protease NAS-37 and anti-protease MLT-11 may therefore be required, respectively, to induce or repress fusion of the seam cells. The gfp fusion gene for the exoribonuclease xrn-2 was expressed in specialized myoepithelial cells that secrete the pharyngeal cuticle (Figures 2D and S1H), consistent with the defect of xrn-2(RNAi) larvae in shedding cuticle from the pharynx. Particular fusion genes were also expressed in specialized interfacial cells that secrete cuticle, including the rectal epithelia (Figure 2A), the excretory duct and pore cells (Figure S1A-B), the vulval epithelium (Figure S1D), and the rectal gland (Figure S1E), as well as support cells for sensory neurons (Figure S1A and S1F).
Figure 2 Expression of Molting Gene gfp Fusion Genes
Expression of GFP (A,C,D,G) or GFP-PEST (B,E,F) from the promoters of the indicated genes.
(A) Fluorescence from qua-1p::gfp in the hypodermis and specialized epithelia.
(B) Fluorescence from nas-37p::gfp-pest in the seam cells and hypodermis of a late L4 stage larva.
(C) Fluorescence from mlt-9p::gfp in the seam cells and hypodermis of a late L3 stage larva.
(D) Fluorescence from xrn-2p::gfp in the pharyngeal myoepithelium (P) of a late L1 stage larva. Only the head of the worm is shown. The less intense fluorescence anterior to the posterior bulb of the pharynx likely corresponds to neurons.
(E) Fluorescence from acn-1p::gfp-pest in the seam cells and hypodermis of a late L1 stage larva.
(F) Fluorescence from mlt-11p::gfp-pest in the seam cells and hypodermis of a late L1 stage larva.
(G) Fluorescence from xrn-2p::gfp in an adult worm, showing the intestine, a neuronal projection along the ventral cord, and a sensory neuron. The anterior of the worm faces right in all pictures
A pulse of fluorescence was observed in the hypodermis prior to molting, for each of the six gfp-pest fusion genes (Figure 3A–C) (unpublished data). Fluorescence from mlt-8p::gfp-pest was first detected approximately 3 h before the L1/L2 molt, or 13 h after hatchlings synchronized by starvation were fed and incubated at 25 °C. The intensity of fluorescence increased until lethargus and then decreased rapidly, such that GFP was barely detectable just 2 h after molting (Figure 3A). When monitoring individual transgenic larvae over the course of development, fluorescence from mlt-8p::gfp-pest was observed from 65 ± 2% to 90 ± 2% of the duration of each larval stage (see Materials and Methods). Expression of mlt-9p::gfp-pest and mlt-10p::gfp-pest in the hypodermis was observed at a similar time, starting, respectively, 64 ± 3% and 63 ± 2% of the way through each larval stage. In contrast, hypodermal expression of GFP from the mlt-11 promoter was detected earlier, from 51 ± 2% to 72 ± 3% of the duration of each stage, suggesting that the MLT-11 anti-protease, synthesized midway through each larval stage, might repress proteases that are post-translationally activated at ecdysis. Expression of mlt-9p::gfp-pest and mlt-11::gfp-pest in the seam cells often preceded and lasted longer than expression in hyp7 (unpublished data). Expression of nas-37p::gfp-pest and acn-1p::gfp-pest also cycled in phase with all four molts (unpublished data). Likewise, expression of qua-1p::gfp in the hypodermis and of xrn-2p::gfp in the pharyngeal myoepithelium intensified prior to molting (unpublished data). Particular fusion genes were also expressed in the epithelial cells of late embryos that synthesize cuticle for the first larval stage. Moreover, expression of the gfp fusion genes was never detected in the hypodermis of gravid adults that no longer molt, whereas mlt-10p::gfp-pest and other fusion genes were expressed in adults that undergo a supernumerary molt due to inactivation of the heterochronic gene lin-29 [62] (unpublished data).
Figure 3 gfp Fusion Genes Are Expressed before Each Molt
(A) Expression of mlt-8p::gfp-pest in an early L1 larva, a larva molting from the L1 to the L2 stage, and an early L2 stage larva. Black arrows indicate cuticle separated from the body of the molting larva. The particular larva shown at L2 was fluorescent before the L1/L2 molt.
(B) Ex[mlt-8p::gfp-pest] (dashed line) or Ex[mlt-10p::gfp-pest] (solid line) larvae were examined for fluorescence and for molting from late in the L1 stage until early adulthood. Graph shows the percent of worms that were fluorescent over time, on a scale normalized to the molting cycle of each worm under observation (see Materials and Methods).
(C) Cycles of fluorescence observed in the hypodermis and seam cells of Ex[mlt-9p::gfp-pest] (dashed line) or Ex[mlt-11p::gfp-pest] (solid line) larvae.
(D) mlt-10 messenger RNA detected by Northern analysis; ribosomal RNA stained with ethidium-bromide provides a loading control.
To verify that cycling fluorescence from a gfp-pest fusion gene reflects dynamic temporal regulation of gene expression, we examined the level of mlt-10 messenger RNA by Northern analysis. As predicted by the mlt-10p::gfp-pest fusion gene, the abundance of mlt-10 mRNA in late L4 larvae exceeded that of mid L4 larvae by a factor of six, and mlt-10 mRNA was barely detectable in young adults (Figure 3D).
Taken together, the spatial and temporal expression patterns of mlt-8, mlt-9, mlt-10, mlt-11, nas-37, acn-1, qua-1, and xrn-2 indicate that the genes are expressed before molting in epithelial cells, such that the corresponding proteins are synthesized in an appropriate time and place to regulate molting. Expression of reporters for mlt-9, mlt-10, mlt-11, and nas-37 in epithelial cells supported our predictions based on gene annotations that the corresponding proteins localize to either the membrane of epithelial cells or the cuticle.
Interestingly, particular fusion genes were expressed in neurons and gland cells that might produce or respond to endocrine signals regulating molting. For example, xrn-2p::gfp was expressed in several anterior neurons, including sensory neurons, as well as the PVT neuron that projects along the ventral cord, and the M5 pharyngeal neuron (Figures 2G, S1G, and S1H). Expression of xrn-2 in the M5 neuron might be relevant to molting because M5 innervates gland cells whose secretions are thought to expedite release of the pharyngeal cuticle [19,35]. Also, xrn-2p::gfp was expressed in M5 only in larvae. The xrn-2 reporter was also expressed in the intestine (Figure 2G), a tissue implicated in the regulation of molting as the site of synthesis of the sterol-modifying enzyme LET-767 [12]. However, expression of xrn-2p::gfp in the intestine persisted in adults that no longer molt, suggesting a function unrelated to molting. The mlt-8 reporter was expressed, in larvae, in a single posterior neuron that remains to be identified. Interestingly, particular fusion genes, such as acn-1p::gfp-pest, were also expressed in the excretory gland cell of larvae (Figure S1A–C). The gland cell is active during ecdysis [63] and is thought to contribute material for the new surface coat. However, ablation of the excretory gland cell does not prevent molting [64], indicating that the essential function of acn-1 in molting is unlikely to stem from expression in this cell. The neurons and other non-epithelial cells that express genes essential for molting represent candidates for foci of endocrine regulation, although the physiologic relevance of gene expression in these non-epithelial tissues remains to be determined. Analysis of the full expression patterns of these and other molting genes awaits the availability of full-length, functional gfp fusion genes that include intronic sequences, or antibodies against the corresponding proteins.
Evidence of an Endocrine Cue for Molting
Observations on the expression of mlt-10p::gfp-pest and other reporters further support the hypothesis of an endocrine cue for C. elegans molting. In nas-37(RNAi) larvae, old cuticle often forms a natural ligature along the longitudinal axis of the worm, typically near the region of the nascent vulva. When Ex[mlt-10p::gfp-pest] nas-37(RNAi) larvae with such ligatures were examined late in the L4 stage, 31% (68/216) expressed GFP exclusively in hypodermis on the anterior side of the constriction, whereas no animals were fluorescent on the posterior side alone (Figure S2). 69% (148/216) of constricted larvae expressed GFP on both sides of the ligature, although in some cases fluorescence was barely detectable on the posterior side. Larvae that expressed GFP only in the anterior section stopped expressing GFP as rapidly as control larvae that completed the last molt, and, in some cases, attempted to shed the L4 cuticle from the head, indicating that the animals were not simply delayed at one point in the molting cycle (unpublished data). Further, larvae that expressed GFP only in the anterior section failed to express GFP in the posterior section up to 8 h after the normal time of the L4-to-adult molt (unpublished data), even though movement often indicated survival of the tissue on the posterior side of the ligature. Together, these observations suggest that expression of mlt-10p::gfp-pest and possibly molting in the posterior hypodermis requires a diffusible cue produced in the anterior of the worm. Similar experiments using man-made ligatures implicated a hormonal cue for molting of the parasitic nematode Aphelenchus avenae in 1967 [65]. Consistent with the view that a cue produced in the anterior of the worm stimulates molting in C. elegans, expression of many of the gfp fusion genes typically begins in the anterior hypodermis and then spreads over time to the anterior and then posterior section of the hypodermal syncytium (hyp7) (unpublished data). Also, the pre-molt cuticle first loosens from around the head during molting [19].
Ordering Gene Expression Cascades Using gfp Fusion Genes
The molting cycle is a complex temporal program likely to have multiple triggers and checkpoints regulating the expression and activity of transcription factors that control multiple downstream targets. We predicted that particular genes isolated in our RNAi screen would act upstream in the molting pathway, directly or indirectly regulating the expression of genes that promote release of the exoskeleton. The availability of GFP reporters allowed us to visualize the gene regulatory status of the pathway when molting was blocked at various points by the inactivation of particular genes required for molting.
To order gene expression cascades among the genes uncovered by RNAi, we fed larvae expressing either the mlt-10 or mlt-8 reporter gene particular dsRNAs of interest. We chose to monitor fluorescence from mlt-10p::gfp-pest because expression of the mlt-10 reporter likely signified the synthesis of components for the new cuticle, and because all of the transgenic animals grew vigorously and expressed GFP late in each larval stage when fed control bacteria. The mlt-8 reporter provided a second marker expressed in the same cells at the same time. We initially examined six gene inactivations representing major functional categories identified in this screen, namely regulators of gene expression (nhr-23), putative signaling pathway components (qua-1, mlt-8, and acn-1), and components of the cuticle or cell membrane (fbn-1 and mlt-9). Transgenic larvae fed the corresponding dsRNAs were monitored for fluorescence and molting over time (Figure 4A). Tracking individual worms ensured that expression of GFP was assessed before larval arrest ensued. Analyzing only animals that failed to molt ensured that a defect in expression of GFP would be detected even if a particular RNAi were effective in the minority of animals fed the bacterial clone.
Figure 4 Ordering Gene Expression Cascades Using gfp Fusion Genes
(A) Ex[mlt-8p::gfp-pest] or Ex[mlt-10p::gfp-pest] larvae were fed bacteria expressing dsRNA for each gene indicated. Graph shows the percent of animals that were fluorescent before a defective molt, normalized to the percent of control larvae that were fluorescent before molting from the same stage. The number of larvae observed is shown in parenthesis. Note that RNAi of mlt-8 or acn-1 typically prevented completion of the L2/L3 molt, whereas RNAi of qua-1, fbn-1, or mlt-9 interfered most often with the L3/L4 or L4/A molts. RNAi of nhr-23 blocked the L3/L4 or L4/A molts in Ex[mlt-10p::gfp-pest] larvae, but prevented completion of the L2/L3 molt in most Ex[mlt-8p::gfp-pest] larvae. In control Ex[mlt-10p::gfp-pest] larvae, fluorescence was observed in 95% (n = 56), 100% (n = 43), or 94% (n = 48) of, respectively, L2, L3, or L4 stage animals. In control Ex[mlt-8p::gfp-pest] larvae, fluorescence was observed in 74% (n = 57) or 70% (n = 36) of L2 stage, and 90% (n = 49) of L4 stage animals. Pair-wise chi-square tests indicate that the decreased fraction of nhr-23(RNAi) or acn-1(RNAi) larvae that express mlt-8p::gfp-pest, and of nhr-23(RNAi), acn-1(RNAi), or mlt-8(RNAi) larvae that express mlt-10p::gfp-pest, relative to control animals, is significant, with p ≤ 0.001 in all cases.
(B) Ex[mlt-10p::gfp-pest] larvae were fed bacteria expressing dsRNA for each gene indicated, or control bacteria not expressing dsRNA of a worm gene. Graph shows the percent of larvae that were fluorescent late in the late L2, L3, and L4 stage, normalized to control animals, with values representing the weighted average of two independent experiments. ‡ indicates that larvae failed to develop to the stage of observation. Table S5 contains the raw data contributing to this figure.
Figure 4A shows that all nhr-23(RNAi) animals failed to express GFP from either the mlt-10 or mlt-8 promoter prior to their ill-fated molt. Inactivation of nhr-23 also diminished expression of the reporters for nas-37, mlt-11, mlt-9, acn-1, and qua-1 in the hypodermis, and of the xrn-2 reporter in the pharyngeal myoepithelium (unpublished data). Thus, NHR-23, synthesized in epithelial cells [13], likely initiates or sustains the pulse of mlt gene expression late in each larval stage, perhaps provoking a response to an as-yet unidentified molting hormone (Figure 5). Inactivation of the acn-1 and mlt-8 genes likewise abrogated expression of GFP from the mlt-10 promoter (Figure 4A), suggesting that ACN-1 and MLT-8 function upstream of mlt-10 but downstream of NHR-23 in this regulatory cascade.
Figure 5 A Model for Molting of C. elegans
(1) Endocrine and possibly neuroendocrine cues trigger molting in C. elegans, stimulating epithelial cells to remodel the exoskeleton near the end of each larval stage. (2) Transcriptional cascades involving NHRs alter gene expression in response to the endocrine cue. In particular, NHR-23 directly or indirectly activates expression of many genes, including mlt-8, mlt-9, mlt-10, mlt-11, acn-1, and nas-37 in the hypodermis, as well as xrn-2 in the pharyngeal myoepithelium. (3) Factors downstream of NHR-23, including MLT-8 and ACN-1, amplify the signal to molt. Signaling via transmembrane proteins likely stimulates release of the old cuticle. (4) Extracellular matrix proteins and secreted enzymes identified in our screen contribute to the new cuticle or regulate release of the old one. We expect precise regulation of these transmembrane proteins and secreted enzymes to accompany the molt. In theory, intercellular signaling might coordinate events in different epithelial cells, the muscle, and the intestine. We further expect secreted signals to provide feedback on the status of the molt to endocrine regulators. The Hint domain protein QUA-1 is a good candidate for a signal secreted from the hypodermis that might amplify a cue for ecdysis, signal to adjacent tissues, or provide feedback. Green shading indicates that a gfp fusion to the corresponding gene was expressed in epithelial cells. † indicates that the gene is required for expression of mlt-10p::gfp-pest in the hypodermis.
In contrast, inactivation of the hedgehog-like gene qua-1, the fibrillin homolog fbn-1, or the MAM domain gene mlt-9 produced larvae that expressed GFP from both the mlt-10 and mlt-8 promoters, but nevertheless failed to complete ecdysis (Figure 4A); indicating that these three genes are dispensable for expression of the mlt-10 and mlt-8 reporters, and likely to function downstream of, or in parallel to, the mlt-10 gene (Figure 5).
To identify additional points of transcriptional control, populations of Ex[mlt-10p::gfp-pest] larvae were fed bacteria expressing dsRNAs corresponding to 76 genes uncovered in our screen, and then monitored for fluorescence late in the L2, L3, and L4 stages. Inactivation of the genes xrn-2, Y65B4A.6, skp-1, D1054.15, R06A4.9, W09B6.1, M03F8.3, T23F2.1, and crs-2, in addition to nhr-23, acn-1, and mlt-8, significantly (p ≤ 0.001) abrogated expression of GFP during a particular stage and blocked development shortly thereafter (Figures 4B and S3; Table S5), suggesting that the corresponding proteins normally induce or sustain expression of mlt-10. The genes identified as putative regulators of the mlt-10 gene encode, respectively, the 5′-3′ exoribonuclease XRN-2, a DEAD (Asp-Glu-Ala-Asp) box helicase, a putative co-factor of NHR-23 [30], two WD-beta repeat proteins, the enzyme acetyl-Coenzyme A carboxylase, a homolog of the spliceosome-associated factor CRN1 [66,67], a glycosyltransferase, and a cysteinyl tRNA synthetase. The xrn-2 gene was verified as a positive regulator of mlt gene expression by tracking fluorescence from individual transgenic larvae over time (unpublished data). Together, we expect the 12 genes upstream of mlt-10p::gfp-pest to be required for epithelial cells to initiate or maintain remodeling of the exoskeleton during molting.
Inactivation of other genes did not significantly (p > 0.001) reduce expression of GFP relative to control larvae of the same stage (Table S5). In most cases, the formal possibility that the corresponding proteins regulate the mlt-10 gene cannot be eliminated due to the variable efficacy of RNAi, particularly if inactivation of a gene partly reduced expression of GFP. However, inactivation of the genes bli-3 (Figure 4B), nas-37, lrp-1, bli-5, ZK430.8, unc-52, lev-11, W10G6.3, kin-2, bli-1, gei-16, K04A8.6, and F25H8.6 produced molting-defective larvae that expressed GFP (unpublished data), demonstrating with certainty that these gene activities are not necessary for expression of mlt-10p::gfp-pest and suggesting, instead, that the genes act downstream of, or in parallel to, the mlt-10 gene. Gene products dispensable for mlt-10 expression might act very near, or, at the time of ecdysis, consistent with our predictions based on gene annotations that the products of fbn-1, bli-3, and ZK430.8 promote assembly or modification of the cuticle. Monitoring expression of the gfp fusion genes thus allowed a first sorting of molting genes uncovered by RNAi into pathways.
Discussion
Using functional genomics, we identified a large set of genes essential for molting in C. elegans. Figure 5 shows a model for the regulation of molting where endocrine or neuroendocrine cues generated by as-yet unidentified cells trigger epithelial cells to remodel the exoskeleton at the end of each larval stage. Particular genes uncovered in this screen encode proteins that regulate gene expression during the molting cycle, whereas other genes encode signaling molecules likely to coordinate the multicellular process of molting. Together, gene annotations as well as spatial and temporal expression studies suggest that many genes identified here specify transmembrane proteins, secreted enzymes, and structural components of the cuticle that are synthesized in epithelial cells and likely regulate the de novo production or release of the exoskeleton at each molt. Thus, activation of the anti-protease MLT-11, the collagenase NAS-37, the MAM domain protein MLT-9, or the LDL-receptor-like protein LRP-1 each represents a potential focus for the spatial and temporal regulation of ecdysis.
Fusions to GFP show that expression of several genes uncovered in this screen cycles in phase with molting, similar to the expression of genes encoding particular NHRs and cuticle collagens [68,69]. Analysis of GFP reporters shows further that NHR-23, directly or indirectly, activates expression of many genes, including mlt-8, mlt-9, mlt-10, mlt-11, nas-37, and acn-1 in the hypodermis, as well as xrn-2 in the pharyngeal myoepithelium. NHR-23 is also required for expression of a cuticle collagen gene, dpy-7 [14], whose product is incorporated into each larval cuticle [69]. Thus, the receptor coordinates gene expression in epithelial cells, possibly in response to an endocrine cue for molting. A ligand for NHR-23 has yet to be identified, but the molecule could be synthesized in neuroendocrine cells or in steroidogenic cells coupled to neurons that regulate molting. Production of a ligand for NHR-23 is likely to be tightly regulated, similar to how steroidogenesis in the prothoracic gland of insects is induced by the neuropeptide prothoracicotropic hormone but also repressed by ecdysteroids [3,70].
Our screen identified the exoribonuclease XRN-2 as a novel regulator of gene expression during molting. One model is that XRN-2 down-regulates the abundance of protein-coding mRNAs or microRNAs that correspond to negative regulators of molting. Together, the observations that xrn-2(RNAi) larvae fail to shed the pharyngeal cuticle and that xrn-2p::gfp is expressed in the pharyngeal myoepithelium suggest that XRN-2 promotes molting in the pharynx. However, xrn-2 is also required for expression of the mlt-10 reporter in the hypodermis, a tissue where expression of xrn-2p::gfp itself has not been detected. One possibility is that XRN-2 activity in the pharynx leads to an intercellular cue that promotes expression of genes in hyp7. Alternatively, xrn-2 might be expressed in the hypodermis, but might not be detectable using this particular gfp fusion gene. XRN-2 and the product of Y65B4A.6, another gene isolated in this screen, might work together to regulate gene expression, because Y65B4A.6 encodes a DEAD-box helicase, and a DEAD-box helicase functions along with the Xrn1p/Rat1p exoribonuclease in mRNA degradation in yeast [71]. Together, the requirement for xrn-2 in molting and for the related gene xrn-1 in embryogenesis [72], establish the XRN family of exoribonucleases as important developmental regulators in C. elegans.
How the multiple genetic pathways uncovered by RNAi converge to regulate gene expression during the molting cycle of C. elegans remains to be determined. However, one or more of these pathways might couple progression of the molting cycle to physiologic or environmental cues including the nutritional status of the larva.
Our identification of putative signaling molecules suggests an essential role for intercellular communication in molting of C. elegans. One idea is that signaling between different epithelia, such as the hypodermal syncytium and the lateral seam cells, might coordinate the production or release of cuticle. Consistent with this view, transcription factors regulating the differentiation and fusion of seam cells are also required for molting [61]. Signaling between the hypodermis and muscle might coordinate remodeling of hemidesmosomes and muscle attachment points. In theory, intercellular signaling might also coordinate division or fusion of the seam cells or endoreduplication of intestinal nuclei with the molt. An alternative view is that cell-autonomous responses to one or a few endocrine cues account for the coordinated activities of different cell types during molting of C. elegans, similar to how different tissues respond to changes in the titer of 20-hydroxyecdysone during insect metamorphosis [2,73].
Particular secreted peptides identified in our screen might amplify endocrine cues for molting. The novel peptide MLT-8 might serve as an autocrine cue that sustains production of the new cuticle, because mlt-8 promotes expression of the GFP reporter for mlt-10 in the same epithelial cells where MLT-8 itself is synthesized. Also, the Hint domain protein QUA-1 is a good candidate for a signal secreted from the hypodermis that could generate a spatially patterned response in the hypodermis itself or in adjacent tissues, thereby coordinating the final stages of the molt.
To set the molting cycle, we expect secreted signals from epithelial cells to provide feedback on the status of the molt to endocrine or neuroendocrine regulators. The existence of physiologic feedback cues is consistent with the observation that many larvae that fail to ecdyse also arrest development, including those larvae defective in proteins that function in epithelial cells, such as LRP-1 [11]. Interference with ecdysone signaling in epidermal tissues similarly triggers a global arrest during Drosophila metamorphosis, suggesting the existence of a molting “checkpoint” in insect development [74]. In theory, any of the signaling components isolated in this screen might function in feedback pathways active during one or more steps of the molt.
Particular peptide hormones might also function in neuroendocrine circuits that regulate the quiescence of larvae during lethargus or the behaviors characteristic of ecdysis, in much the same way that peptide hormones of insects trigger behaviors essential for escape from the old exoskeleton [4,8]. However, none of the putative secreted peptides identified in this screen show obvious sequence similarity to eclosion hormone or ecdysis-triggering hormone.
Genes or hormones that function far upstream in the molting pathway of nematodes can now be identified, respectively, as mutations or compounds that alter the timing of expression of the cycling GFP reporters. Master regulators of molting in nematodes might function in endocrine or possibly neuroendocrine cells and might be conserved in arthropods, given that molting is a universal feature of the Ecdysozoan clade [1]. One simple explanation for the abundance of epithelial, as opposed to neuronal, genes uncovered in the screen described here is that RNAi works better in epithelial cells than in neurons [75]. Particular gene inactivations that produced molting defects at low penetrance in our screen (Table S4) might therefore correspond to endocrine or neuroendocrine components. In addition, a screen for arrest during ecdysis, rather than a screen for aberrant timing of the molt, might enrich for epithelial factors.
Identifying genes essential for molting of C. elegans enables the development of safe and effective insecticides and nematicides that target gene products conserved only in Ecdysozoans. Molting genes conserved only in insects and nematodes, such as the extracellular matrix proteins NOAH-1 and NOAH-2, identify potential targets for insecticides expected to harm only Ecdysozoans. Current anti-nematode drugs, such as benzimidazoles and avermectins, target, respectively, cytoskeletal components and ion channels that are conserved in mammals, and the drugs therefore can be toxic to humans. Resistance to these compounds is also increasingly common [76,77]. One potential new drug target is MLT-8, since the corresponding gene is conserved and highly expressed at the molt in a parasitic nematode, as inferred by the identification of 32 cDNAs matching C. elegans mlt-8 (p = E-121) in a library derived from molting O. volvulus (Table S7) (unpublished data). However, the novelty of MLT-8 may pose a considerable challenge for drug development. In this regard, molting proteases, like NAS-37, represent more attractive targets for the development of small-molecule antagonists, given the success of drug development on protease targets for high blood pressure and HIV [78,79].
Materials and Methods
Screening the RNAi library for molting genes
Approximately 16,757 bacterial clones that each express a 0.5- to 2-kilobase dsRNA corresponding to one worm gene were obtained courtesy of J. Ahringer's laboratory [26,27]. An additional 1,821 bacterial clones that express dsRNAs corresponding to worm genes not represented in the Ahringer library were obtained courtesy of M. Vidal [28]. Bacterial clones expressing dsRNA of worm genes were cultured as described [26], except that nematode growth medium was supplemented with 8 mM IPTG and 25 μg/ml carbenicillin. Approximately 25 wild-type (N2) hatchlings (early L1 larvae) were isolated using standard techniques, fed a particular bacterial clone, and cultivated for 2.5 d at 20 °C before visual inspection. Although animals were examined blind to the identity of the dsRNA, the library designation for each bacterial clone could be readily decoded to reveal the corresponding gene [27]. Gene inactivations observed to cause molting defects in the primary library screen were tested again by feeding the bacterial clones to about 50 wild-type (N2) or 50 rrf-3(pk1426) mutant larvae [29]. In a control, larvae with molting defects were not observed after about 1,000 N2 or rrf-3(pk1426) animals were fed isogenic bacteria not expressing dsRNA of worm genes. Also, the vast majority of E. coli strains expressing dsRNAs of particular worm genes caused no molting defects. To verify the identity of genes inactivated by RNAi, plasmid DNA was isolated from each bacterial clone of interest and the insert DNA sequenced using primers complementary to the vector pPD129.36 [80]. In this screen, 7% of clones contained worm DNA different from the expected insert. Sequence and protein names refer to designations by WormBase (http://www.wormbase.org/).
To evaluate the dauer molt, approximately 20 hatchlings of the temperature-sensitive, dauer-constitutive mutants daf-2(e1370) [81] and daf-7(e1372) [82] were fed bacterial clones expressing dsRNA for each molting gene, in duplicate, and then cultivated at restrictive temperature (25 °C) for 3 d, such that control animals all became dauers. Animals were then shifted to permissive temperature (15 °C) for 2 d, allowing control animals to molt to the L4 stage. Observation of L2d or dauer larvae with the Mlt phenotype, in either genetic background, indicated that a given gene inactivation disrupted the L2d/dauer or dauer/L4 molt (Table S6).
Strains and molecular constructs
Strains used for this study appear in Table S8. Table S9 shows the PCR primers used to construct gfp fusion genes. To append the PEST sequence to the C-terminus of GFP, nucleotides 1,399–1,521 of pd1EGFP-N1 (Clontech, Palo Alto, California, United States) were inserted into pPD95_81 (A. Fire) between the last coding codon and the stop codon of gfp, generating pAF207. For each gene, primers U1 and FL were used to amplify N2 genomic DNA corresponding to the initiation codon and upstream sequence, while primers FU and CAW31 (5′-
GCCGCATAGTTAAGCCAGCC 3′) [83] were used to amplify DNA corresponding to the gfp-pest or gfp gene and the 3′ UTR of unc-54 from, respectively, pAF207 or pPD95_81. The PCR products were annealed and the resulting polynucleotide amplified using primers U2 and CAW32 (5′
CCGCTTACAGACAAG
CTGTGA 3′) under conditions described previously [83]. To generate the extrachromosomal arrays mgEx646, mgEx647, mgEx648, mgEx649, mgEx656, mgEx654, mgEx650, mgEx675, and mgEx676, PCR products corresponding to, respectively, mlt-10p::gfp-pest, mlt-8p::gfp-pest, mlt-9p::gfp-pest, mlt-11p::gfp-pest, nas-37p::gfp-pest, acn-1p::gfp-pest, mlt-9p::gfp, qua-1p::gfp, and xrn-2p::gfp were microinjected at 5 to 10 ng/μl into temperature-sensitive pha-1(e2123) mutant animals along with the pha-1(+) plasmid pBX [84] at 3 ng/μl and pBS DNA to a final concentration of 100 ng/μl. Use of the pha-1(e2123) genetic background allowed for the recovery and cultivation of worm populations in which virtually all of the animals maintained the extrachromosomal array, because only pha-1(+) transgenic embryos survive at 25 °C [84]. Verifying that GFP-PEST fusion proteins are degraded by the proteosome, we found that RNAi of proteosome subunit genes like pbs-5 prolonged fluorescence from mlt-10p::gfp-pest even in developmentally arrested larvae (Table S5 and Figure S3). Fusions between the conventional gfp gene and the promoters of qua-1 and xrn-2 were used for expression studies because fluorescence was not readily detected from fusions to the gfp-pest gene.
Comparative sequence searches
Sequence homologs of the MLT proteins were identified using the standard algorithm TBlastN [85] to compare the predicted product of each molting gene to the GenBank library of translated DNA sequences from H. sapiens, D. melanogaster, and S. cerevisiae, or translated cDNA sequences from nematodes. The most highly related proteins from human, fly, or yeast were then compared against all worm proteins using the algorithm BlastX and the WormPep database at WormBase (http://www.wormbase.org). Signal peptides were identified in predicted gene products using either the SignalP 3.0 server (http://www.cbs.dtu.dk/services/SignalP/) or Kyte-Doolittle Hydropathy plots.
Monitoring gfp fusion gene expression
To monitor temporal expression of the mlt gene gfp-pest fusion genes, synchronized L1 hatchlings of GR1348, GR1349, GR1350, or GR1351 (Table S8) were plated on nematode growth medium with E. coli strain OP50 as a food source and incubated at 25 °C. Fluorescent larvae were selected 14 h later to ensure the use of non-mosaic, highly synchronous animals. Each larva was scored every hour for detectable fluorescence, using a Zeiss (Oberkochen, Germany) Stemi-SV6 microscope, and for molting, indicated by shedding of the cuticle. Each animal was transferred to a new plate after each molt. In Figure 3, we report the percent of animals that were fluorescent over time, on a scale normalized to the period between molts for each worm under observation. As an example, a larva that molted from L1 to L2 at noon, molted from L2 to L3 at 8 P.M., was fluorescent at 7 P.M. and 8 P.M., but was not fluorescent at 6 P.M. or 9 P.M., would be recorded as fluorescent from time 1.87 to time 2.0, or, from 87.5% to 100% of the L2 stage. Calculations of the average duration of fluorescence, reported with the 95% confidence interval, include observations from larvae during the L2, L3, and L4 stages. Because many of the extrachromosomal arrays were associated with some larval lethality, only larvae that completed all four molts were included in the final analysis. Twenty-four larvae were observed for mlt-8p::gfp-pest whereas 20 larvae were observed for all other reporters. For northern analysis, RNA from extracts of mid L4, late L4, and young adult animals was resolved and hybridized with a mlt-10 probe, corresponding to base pairs 5,070 to 6,997 of cosmid C09E8, as previously described [86]. Message levels were quantified using ImageQuant software and a phosphorimager.
To order gene expression cascades, synchronized hatchlings of mgEx646[mlt-10p::gfp-pest] (GR1348) and mgEx647[mlt-8p::gfp-pest] (GR1349) were fed bacteria expressing dsRNA for each gene of interest, or, as a control, fed isogenic bacteria not expressing dsRNA for a worm gene. After incubation for no more than 15 h at 25 °C, a fluorescent larva was transferred to each well of several 24-well RNAi plates seeded with the appropriate bacteria. For each developmental stage, larvae were observed over a 6 to 9 h time period starting when control larvae first became fluorescent, and scored every 2–3 h for detectable fluorescence and for the Mlt phenotype. In Figure 4A, we report the percent of animals that were fluorescent prior to a defective molt, normalized to the fraction of control larvae that were fluorescent before molting from the same stage. To screen the full set of molting gene inactivations, approximately 20 synchronized hatchlings of GR1348 were fed bacteria expressing dsRNA corresponding to each gene of interest in two trials. The percent of larvae with detectable fluorescence was scored 1–3 h before control larvae molted from the L2, L3, or L4 stage, when the majority of control larvae were fluorescent. Only larvae of the same developmental stage as control animals were scored at each time point. To examine patterns of GFP expression in nas-37(RNAi) mgEx646[mlt10p::gfp-pest] larvae, hatchlings of GR1348 were fed bacteria expressing dsRNA for nas-37 and L4 stage larvae examined after 39–43 h of growth at 25 °C. Larvae with coronal constrictions were scored for fluorescence using a Zeiss M2-Bio microscope. Twenty-six larvae that expressed GFP only anterior of the ligature, and 22 fluorescent control larvae, were then transferred to new plates and each worm was observed for fluorescence and molting 4 and 8 h later.
Supporting Information
Figure S1 Expression of Molting Gene gfp Fusion Genes
(A–C) Expression of acn-1p::gfp-pest. (A) Fluorescence in the excretory gland, duct, and pore cells (Exc), and in the glial cells (Gi) of interlabial neurons. (B) Nomarski image of the same larva. (C) Expression in the excretory gland cell (Gn). (D and E) Expression of nas-37p::gfp-pest in specialized epithelia. (D) Expression in the vulva. (E) Fluorescence in the rectal gland, labeled RG. The solid line traces the tail of the worm, and the dashed line shows the edges of the intestine. (F) Expression of qua-1p::gfp in support cells for head neurons. (G and H) Expression of xrn-2p::gfp. Expression in the M5 pharyngeal neuron (G) and posterior bulb of the pharynx (H) of the same larva. (C), (D), and (F) show fluorescence images superimposed on Nomarski optics. P, posterior bulb of the pharynx.
(26 MB EPS).
Click here for additional data file.
Figure S2 Effect of Ligatures on Expression of mlt-10p::gfp-pest
Patterns of GFP expression in nas-37(RNAi) Ex[mlt10p::gfp-pest] larvae observed at the L4 stage. Diagrams show the anterior (A) of the worm facing left, and green color indicates expression of GFP. 216 larvae selected for observation at the L4 stage were constricted by old cuticle from a prior larval stage. The number of larvae observed in each class is shown in parenthesis. Approximately 7% (n = 1,109) of animals lacking a ligature showed more intense expression of GFP in hypodermal tissues anterior of the nascent vulva at the time of observation.
(3.2 MB EPS).
Click here for additional data file.
Figure S3 Expression of mlt-10p::gfp-pest After RNAi of Genes Important for Molting
Ex[mlt-10p::gfp-pest] larvae were fed bacteria expressing dsRNA for each gene indicated, or control bacteria not expressing dsRNA of a worm gene. Graph shows the percent of larvae that were fluorescent in the late L4 stage, normalized to control animals. Values represent the weighted average of two independent trials, with an average of 38 larvae examined per bacterial clone. Asterisks indicate trials where the fraction of fluorescent larvae differed significantly from that of controls (p ≤ 0.001 in pair-wise chi-square tests). Table S5 contains the raw data contributing to this figure.
(735 KB EPS).
Click here for additional data file.
Table S1 Genes Whose Inactivation Disrupts Molting and Their Relation to Genes in Other Species
(1) Sequence names as designated by WormBase (http://www.wormbase.org/)
(2) Top hits from TBlastN searches of the human or fly genome using the predicted C. elegans gene product. Red shading of the text indicates that a BlastX search with the predicted human or fly protein uncovered the corresponding C. elegans protein as the top-scoring match in C. elegans, suggesting homology.
(56 KB XLS).
Click here for additional data file.
Table S2 Genes Required for General Secretion and Endocytosis Uncovered in the Molting Screen
(19 KB XLS).
Click here for additional data file.
Table S3 Genes Required for Protein Synthesis Uncovered in the Molting Screen
(20 KB XLS).
Click here for additional data file.
Table S4 Gene Inactivations Producing Molting Defects in 10% or Less of Larvae
(23 KB XLS).
Click here for additional data file.
Table S5 Expression of mlt-10p::gfp-pest during RNAi of Molting Genes
* Vector B is the control sample for RNAi of K04A8.6, ZC13.3, ZK945.2, C15H11.7, F32D8.6, T19A5.3, Y65B4A.6, Y23H5A.1, R06A4.9, Y47D3B.1, Y54E10BR.5, T25B9.9, T17H7.4, Y105E8B.1, W03F9.10, F56C9.1, C23G10.10, and F25H8.6. Vector A is the control for all other gene inactivations.
N.A., not applicable (RNAi caused larval arrest at an earlier stage); N.D., not determined; #, RNAi produced larvae arrested at the L2 to L3 stage that continued to express GFP at this time point.
(53 KB XLS).
Click here for additional data file.
Table S6 Inactivation of mlt Genes in Dauer Larvae
+ indicates the observation of dauer larvae trapped in cuticle among approximately 40 of the dauer-constitutive conditional mutants daf-2(e1370) or daf-7(e1372). Hatchlings were incubated at 25 °C to drive dauer formation, and then at 15 °C to allow recovery and a molt to the L4 stage. Note that dsRNAs that cause a larval arrest at the L1 or L2 stage (Table S5) could not be evaluated using this method.
(22 KB XLS).
Click here for additional data file.
Table S7 Homologs of Selected Molting Genes in Parasitic Nematodes
1 Top hits and scores from TBlastN searches with the predicted C. elegans gene product versus translated cDNAs isolated from the indicated species.
(79 KB XLS).
Click here for additional data file.
Table S8 Strains Used in This Study
(17 KB XLS).
Click here for additional data file.
Table S9 Primers Used for Construction of gfp Fusion Genes
R1 refers to the sequence 5′-
CGGGATTGGCCAAAGGACCCAAAG-3′; R2 refers to the sequence 5′-
CTTTGGGTCCTTTGGCCAATCCCG-3′
(19 KB XLS).
Click here for additional data file.
Accession Numbers
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank) accession number for cosmid C09E8 is AF077529.
The WormBase (http://www.wormbase.org/) accession numbers for genes and gene products discussed in this paper are acn-1 (C42D8.5), bli-1 (C09G5.6), BLI-3 (WP:CE28463), bli-4 (K04F10.4), bli-5 (F45G2.5), crs-2 (Y23H5A.1), DAB-1 (WP:CE36446), dpy-7 (F46C8.6), fbn-1 (ZK783.1), gei-16 (T17H7.4), kin-2 (R07E4.6), LET-767 (WP:CE30639), lev-11 (Y105E8B.1), LIN-26 (WP:CE27972), lin-29 (W03C9.4), lir-1 (F18A1.3), lrp-1 (F29D11.1), LRP-1 (WP:CE05765), mlt-8 (W08F4.6), mlt-9 (F09B12.1), mlt-10 (C09E8.3), mlt-11(W01F3.3), nas-36 (C26C6.3), nas-37 (C17G1.6), nhr-23 (C01H6.5), NHR-23 (WP:CE24775), nhr-25 (F11C1.6), NHR-25 (WP:CE03191), noah-1 (C34G6.6), noah-2 (F52B11.3), PAN-1 (WP:CE01041), PEB-1 (WP:CE07500), ptr-23 (ZK270.1), ptr-4 (C45B2.7), qua-1 (T05C12.10), rme-8 (F18C12.2), skp-1 (T27F2.1), unc-52 (ZC101.2), xrn-1 (Y39G8C.1), and xrn-2 (Y48B6A.3).
We are grateful to Julie Ahringer for creating and sharing the bacterial RNAi library prior to publication. We thank J. F. Rual, J. Ceron, J. Koreth, T. Hao, A. S. Nicot, T. Hirozane-Kishikawa, J. Vandenhaute, S. H. Orkin, D. E. Hill, S. van den Heuval, and Marc Vidal for creating and sharing additional bacterial RNAi clones, and John Kim, Ravi Kamath, and Harrison Gabel for organizing bacterial clones from the Vidal library. Snjezana Joksimovic and Jinling Xu provided technical assistance. We are grateful to John Yochem for critical reading of the manuscript and valuable discussions. We thank all members of the Ruvkun lab for their support, both scientific and personal. Some strains used in this study were provided by the Caenorhabditis Genetics Center (CGC). This work was supported by a Jane Coffin Childs Memorial Foundation post-doctoral research fellowship to ARF, a Massachusetts General Hospital Fund for Medical Discovery post-doctoral research fellowship to ARF, and a National Institutes of Health grant to GR.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. ARF and GR conceived and designed the experiments. ARF and SR performed the experiments. ARF wrote the paper.
Citation: Frand AR, Russel S, Ruvkun G (2005) Functional genomic analysis of C. elegans molting. PLoS Biol 3(10): e312
Abbreviations
ACEangiotensin converting enzyme
GFPgreen fluorescent protein
LDLlow-density lipoprotein
NHRnuclear hormone receptor
RNAiRNA-interference
==== Refs
References
Aguinaldo AM Turbeville JM Linford LS Rivera MC Garey JR Evidence for a clade of nematodes, arthropods and other moulting animals Nature 1997 387 489 493 9168109
Thummel CS Files on steroids—Drosophila metamorphosis and the mechanisms of steroid hormone action Trends Genet 1996 12 306 310 8783940
Gilbert LI Rybczynski R Warren JT Control and biochemical nature of the ecdysteroidogenic pathway Annu Rev Entomol 2002 47 883 916 11729094
Truman JW The eclosion hormone system of insects Prog Brain Res 1992 92 361 374 1302884
Zitnan D Ross LS Zitnanova I Hermesman JL Gill SS Steroid induction of a peptide hormone gene leads to orchestration of a defined behavioral sequence Neuron 1999 23 523 535 10433264
Kingan TG Gray W Zitnan D Adams ME Regulation of ecdysis-triggering hormone release by eclosion hormone J Exp Biol 1997 200 3245 3256 9364030
Kingan TG Adams ME Ecdysteroids regulate secretory competence in Inka cells J Exp Biol 2000 203 3011 3018 10976037
Zitnan D Kingan TG Hermesman JL Adams ME Identification of ecdysis-triggering hormone from an epitracheal endocrine system Science 1996 271 88 91 8539606
Park Y Filippov V Gill SS Adams ME Deletion of the ecdysis-triggering hormone gene leads to lethal ecdysis deficiency Development 2002 129 493 503 11807040
Ewer J Gammie SC Truman JW Control of insect ecdysis by a positive-feedback endocrine system: Roles of eclosion hormone and ecdysis triggering hormone J Exp Biol 1997 200 869 881 9100362
Yochem J Tuck S Greenwald I Han M A gp330/megalin-related protein is required in the major epidermis of Caenorhabditis elegans for completion of molting Development 1999 126 597 606 9876188
Kuervers LM Jones CL O'Neil NJ Baillie DL The sterol modifying enzyme LET-767 is essential for growth, reproduction and development in Caenorhabditis elegans
Mol Genet Genomics 2003 270 121 131 12905072
Kostrouchova M Krause M Kostrouch Z Rall JE CHR3: A Caenorhabditis elegans orphan nuclear hormone receptor required for proper epidermal development and molting Development 1998 125 1617 1626 9521900
Kostrouchova M Krause M Kostrouch Z Rall JE Nuclear hormone receptor CHR3 is a critical regulator of all four larval molts of the nematode Caenorhabditis elegans
Proc Natl Acad Sci U S A 2001 98 7360 7365 11416209
Gissendanner CR Sluder AE nhr-25, the Caenorhabditis elegans ortholog of ftz-f1, is required for epidermal and somatic gonad development Dev Biol 2000 221 259 272 10772806
Asahina M Ishihara T Jindra M Kohara Y Katsura I The conserved nuclear receptor Ftz-F1 is required for embryogenesis, moulting and reproduction in Caenorhabditis elegans
Genes Cells 2000 5 711 723 10971653
Chitwood DJ Biochemistry and function of nematode steroids Crit Rev Biochem Mol Biol 1999 34 273 284 10517647
Sluder AE Maina CV Nuclear receptors in nematodes: Themes and variations Trends Genet 2001 17 206 213 11275326
Singh RN Soulston JE Some observations on moulting in Caenorhabditis elegans
Nematologica 1978 24 63 71
Johnstone IL The cuticle of the nematode Caenorhabditis elegans A complex collagen structure Bioessays 1994 16 171 178 8166670
Nelson FK Albert PS Riddle DL Fine structure of the Caenorhabditis elegans secretory-excretory system J Ultrastruct Res 1983 82 156 171 6827646
Knight CG Patel MN Azevedo RB Leroi AM A novel mode of ecdysozoan growth in Caenorhabditis elegans
Evol Dev 2002 4 16 27 11871396
Davis MW Birnie AJ Chan AC Page AP Jorgensen EM A conserved metalloprotease mediates ecdysis in Caenorhabditis elegans
Development 2004 131 6001 6008 15539494
Suzuki M Sagoh N Iwasaki H Inoue H Takahashi K Metalloproteases with EGF, CUB, and thrombospondin-1 domains function in molting of Caenorhabditis elegans
Biol Chem 2004 385 565 568 15255192
Richards FO Boatin B Sauerbrey M Seketeli A Control of onchocerciasis today: Status and challenges Trends Parasitol 2001 17 558 563 11756018
Fraser AG Kamath RS Zipperlen P Martinez-Campos M Sohrmann M Functional genomic analysis of C. elegans chromosome I by systematic RNA interference Nature 2000 408 325 330 11099033
Kamath RS Fraser AG Dong Y Poulin G Durbin R Systematic functional analysis of the Caenorhabditis elegans genome using RNAi Nature 2003 421 231 237 12529635
Rual JF Ceron J Koreth J Hao T Nicot AS Toward improving Caenorhabditis elegans phenome mapping with an ORFeome-based RNAi library Genome Res 2004 14 2162 2168 15489339
Simmer F Tijsterman M Parrish S Koushika SP Nonet ML Loss of the putative RNA-directed RNA polymerase RRF-3 makes C. elegans hypersensitive to RNAi Curr Biol 2002 12 1317 1319 12176360
Kostrouchova M Housa D Kostrouch Z Saudek V Rall JE SKIP is an indispensable factor for Caenorhabditis elegans development Proc Natl Acad Sci U S A 2002 99 9254 9259 12084813
Zhang Y Grant B Hirsh D RME-8, a conserved J-domain protein, is required for endocytosis in Caenorhabditis elegans
Mol Biol Cell 2001 12 2011 2021 11451999
Brooks DR Appleford PJ Murray L Isaac RE An essential role in molting and morphogenesis of Caenorhabditis elegans for ACN-1, a novel member of the angiotensin-converting enzyme family that lacks a metallopeptidase active site J Biol Chem 2003 278 52340 52346 14559923
Peters K McDowall J Rose AM Mutations in the bli-4 (I) locus of Caenorhabditis elegans disrupt both adult cuticle and early larval development Genetics 1991 129 95 102 1936966
Simmer F Moorman C van der Linden AM Kuijk E van den Berghe PV Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions PLoS Biol 2003 1 e12 14551910
Fernandez AP Gibbons J Okkema PG
C. elegans peb-1 mutants exhibit pleiotropic defects in molting, feeding, and morphology Dev Biol 2004 276 352 366 15581870
Li T Bender M A conditional rescue system reveals essential functions for the ecdysone receptor (EcR) gene during molting and metamorphosis in Drosophila
Development 2000 127 2897 2905 10851134
Dufourcq P Chanal P Vicaire S Camut E Quintin S lir-2, lir-1 and lin-26 encode a new class of zinc-finger proteins and are organized in two overlapping operons both in Caenorhabditis elegans and in Caenorhabditis briggsae
Genetics 1999 152 221 235 10224256
Zhang M Yu L Xin Y Hu P Fu Q Cloning and mapping of the XRN2 gene to human chromosome 20p11.1-p11.2 Genomics 1999 59 252 254 10409438
Petfalski E Dandekar T Henry Y Tollervey D Processing of the precursors to small nucleolar RNAs and rRNAs requires common components Mol Cell Biol 1998 18 1181 1189 9488433
Das B Butler JS Sherman F Degradation of normal mRNA in the nucleus of Saccharomyces cerevisiae
Mol Cell Biol 2003 23 5502 5515 12897126
Bousquet-Antonelli C Presutti C Tollervey D Identification of a regulated pathway for nuclear pre-mRNA turnover Cell 2000 102 765 775 11030620
Aspock G Kagoshima H Niklaus G Burglin TR
Caenorhabditis elegans has scores of hedgehog-related genes: Sequence and expression analysis Genome Res 1999 9 909 923 10523520
Lum L Beachy PA The Hedgehog response network: Sensors, switches, and routers Science 2004 304 1755 1759 15205520
Barlowe C Traffic COPs of the early secretory pathway Traffic 2000 1 371 377 11208122
Roberts B Clucas C Johnstone IL Loss of SEC-23 in Caenorhabditis elegans causes defects in oogenesis, morphogenesis, and extracellular matrix secretion Mol Biol Cell 2003 14 4414 4426 14551256
Kamikura DM Cooper JA Lipoprotein receptors and a disabled family cytoplasmic adaptor protein regulate EGL-17/FGF export in C. elegans
Genes Dev 2003 17 2798 2811 14630941
Loeys BL Matthys DM de Paepe AM Genetic fibrillinopathies: New insights in molecular diagnosis and clinical management Acta Clin Belg 2003 58 3 11 12723256
Chung YD Zhu J Han Y Kernan MJ nompA encodes a PNS-specific, ZP domain protein required to connect mechanosensory dendrites to sensory structures Neuron 2001 29 415 428 11239432
Edens WA Sharling L Cheng G Shapira R Kinkade JM Tyrosine cross-linking of extracellular matrix is catalyzed by Duox, a multidomain oxidase/peroxidase with homology to the phagocyte oxidase subunit gp91phox J Cell Biol 2001 154 879 891 11514595
Unsold C Pappano WN Imamura Y Steiglitz BM Greenspan DS Biosynthetic processing of the pro-alpha 1(V)2pro-alpha 2(V) collagen heterotrimer by bone morphogenetic protein-1 and furin-like proprotein convertases J Biol Chem 2002 277 5596 5602 11741999
Rattenholl A Pappano WN Koch M Keene DR Kadler KE Proteinases of the bone morphogenetic protein-1 family convert procollagen VII to mature anchoring fibril collagen J Biol Chem 2002 277 26372 26378 11986329
Kramerova IA Kawaguchi N Fessler LI Nelson RE Chen Y Papilin in development; A pericellular protein with a homology to the ADAMTS metalloproteinases Development 2000 127 5475 5485 11076767
Beckmann G Bork P An adhesive domain detected in functionally diverse receptors Trends Biochem Sci 1993 18 40 41 8387703
Grigorenko AP Moliaka YK Soto MC Mello CC Rogaev EI The Caenorhabditis elegans IMPAS gene, imp-2, is essential for development and is functionally distinct from related presenilins Proc Natl Acad Sci U S A 2004 101 14955 14960 15469912
Hresko MC Schriefer LA Shrimankar P Waterston RH Myotactin, a novel hypodermal protein involved in muscle-cell adhesion in Caenorhabditis elegans
J Cell Biol 1999 146 659 672 10444073
Hong L Elbl T Ward J Franzini-Armstrong C Rybicka KK MUP-4 is a novel transmembrane protein with functions in epithelial cell adhesion in Caenorhabditis elegans
J Cell Biol 2001 154 403 414 11470827
Williams BD Waterston RH Genes critical for muscle development and function in Caenorhabditis elegans identified through lethal mutations J Cell Biol 1994 124 475 490 8106547
Rogalski TM Williams BD Mullen GP Moerman DG Products of the unc-52 gene in Caenorhabditis elegans are homologous to the core protein of the mammalian basement membrane heparan sulfate proteoglycan Genes Dev 1993 7 1471 1484 8393416
Hresko MC Williams BD Waterston RH Assembly of body wall muscle and muscle cell attachment structures in Caenorhabditis elegans
J Cell Biol 1994 124 491 506 8106548
Li X Zhao X Fang Y Jiang X Duong T Generation of destabilized green fluorescent protein as a transcription reporter J Biol Chem 1998 273 34970 34975 9857028
Koh K Rothman JH ELT-5 and ELT-6 are required continuously to regulate epidermal seam cell differentiation and cell fusion in C. elegans
Development 2001 128 2867 2880 11532911
Papp A Rougvie AE Ambros V Molecular cloning of lin-29, a heterochronic gene required for the differentiation of hypodermal cells and the cessation of molting in C. elegans
Nucleic Acids Res 1991 19 623 630 1672752
Chitwood BG Chitwood MB An introduction to nematology 1950 Baltimore University Park Press 334
Nelson FK Riddle DL Functional study of the Caenorhabditis elegans secretory-excretory system using laser microsurgery J Exp Zool 1984 231 45 56 6470649
Davies KA Fisher JM On hormonal control of moulting in Aphelenchus avenae (Nematoda: Aphelenchida) Int J Parasitol 1994 24 649 655 7928066
Chung S Zhou Z Huddleston KA Harrison DA Reed R Crooked neck is a component of the human spliceosome and implicated in the splicing process Biochim Biophys Acta 2002 1576 287 297 12084575
Raisin-Tani S Leopold P
Drosophila crooked-neck protein co-fractionates in a multiprotein complex with splicing factors Biochem Biophys Res Commun 2002 296 288 292 12163015
Gissendanner CR Crossgrove K Kraus KA Maina CV Sluder AE Expression and function of conserved nuclear receptor genes in Caenorhabditis elegans
Dev Biol 2004 266 399 416 14738886
Johnstone IL Barry JD Temporal reiteration of a precise gene expression pattern during nematode development Embo J 1996 15 3633 3639 8670866
Sakurai S Williams CM Short-loop negative and positive feedback on ecdysone secretion by prothoracic gland in the tobacco hornworm, Manduca sexta
Gen Comp Endocrinol 1989 75 204 216 2806870
Bond AT Mangus DA He F Jacobson A Absence of Dbp2p alters both nonsense-mediated mRNA decay and rRNA processing Mol Cell Biol 2001 21 7366 7379 11585918
Newbury S Woollard A The 5′-3′ exoribonuclease xrn-1 is essential for ventral epithelial enclosure during C. elegans embryogenesis RNA 2004 10 59 65 14681585
Lee T Marticke S Sung C Robinow S Luo L Cell-autonomous requirement of the USP/EcR-B ecdysone receptor for mushroom body neuronal remodeling in Drosophila
Neuron 2000 28 807 818 11163268
Cherbas L Hu X Zhimulev I Belyaeva E Cherbas P EcR isoforms in Drosophila : Testing tissue-specific requirements by targeted blockade and rescue Development 2003 130 271 284 12466195
Timmons L Court DL Fire A Ingestion of bacterially expressed dsRNAs can produce specific and potent genetic interference in Caenorhabditis elegans
Gene 2001 263 103 112 11223248
Dent JA Smith MM Vassilatis DK Avery L The genetics of ivermectin resistance in Caenorhabditis elegans
Proc Natl Acad Sci U S A 2000 97 2674 2679 10716995
Behm CA Bendig MM McCarter JP Sluder AE RNAi-based discovery and validation of new drug targets in filarial nematodes Trends Parasitol 2005 21 97 100 15734653
Vane JR The history of inhibitors of angiotensin converting enzyme J Physiol Pharmacol 1999 50 489 498 10639000
Cvetkovic RS Goa KL Lopinavir/ritonavir: A review of its use in the management of HIV infection Drugs 2003 63 769 802 12662125
Timmons L Fire A Specific interference by ingested dsRNA Nature 1998 395 854 9804418
Kimura KD Tissenbaum HA Liu Y Ruvkun G daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans
Science 1997 277 942 946 9252323
Ren P Lim CS Johnsen R Albert PS Pilgrim D Control of C. elegans larval development by neuronal expression of a TGF-beta homolog Science 1996 274 1389 1391 8910282
Wolkow CA Kimura KD Lee MS Ruvkun G Regulation of C. elegans life-span by insulinlike signaling in the nervous system Science 2000 290 147 150 11021802
Granato M Schnabel H Schnabel R pha-1, a selectable marker for gene transfer in C. elegans
Nucleic Acids Res 1994 22 1762 1763 8202383
Altschul SF Gish W Miller W Myers EW Lipman DJ Basic local alignment search tool J Mol Biol 1990 215 403 410 2231712
Lee SS Kennedy S Tolonen AC Ruvkun G DAF-16 target genes that control C. elegans life-span and metabolism Science 2003 300 644 647 12690206
|
16122351
|
PMC1233573
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Oct 30; 3(10):e312
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030312
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1612234810.1371/journal.pbio.0030316Research ArticleBioinformatics/Computational BiologyEvolutionGenetics/Genomics/Gene TherapyEubacteriaArchaeaThe Cobweb of Life Revealed by Genome-Scale Estimates of Horizontal Gene Transfer Genome-Scale Horizontal Gene TransferGe Fan
1
Wang Li-San
1
Kim Junhyong [email protected]
1
1Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania, United States of AmericaHillis David Academic EditorUniversity of TexasUnited States of America10 2005 30 8 2005 30 8 2005 3 10 e31624 2 2005 11 7 2005 Copyright: © 2005 Ge 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.
Comparing Gene Trees and Genome Trees: A Cobweb of Life?
With the availability of increasing amounts of genomic sequences, it is becoming clear that genomes experience horizontal transfer and incorporation of genetic information. However, to what extent such horizontal gene transfer (HGT) affects the core genealogical history of organisms remains controversial. Based on initial analyses of complete genomic sequences, HGT has been suggested to be so widespread that it might be the “essence of phylogeny” and might leave the treelike form of genealogy in doubt. On the other hand, possible biased estimation of HGT extent and the findings of coherent phylogenetic patterns indicate that phylogeny of life is well represented by tree graphs. Here, we reexamine this question by assessing the extent of HGT among core orthologous genes using a novel statistical method based on statistical comparisons of tree topology. We apply the method to 40 microbial genomes in the Clusters of Orthologous Groups database over a curated set of 297 orthologous gene clusters, and we detect significant HGT events in 33 out of 297 clusters over a wide range of functional categories. Estimates of positions of HGT events suggest a low mean genome-specific rate of HGT (2.0%) among the orthologous genes, which is in general agreement with other quantitative of HGT. We propose that HGT events, even when relatively common, still leave the treelike history of phylogenies intact, much like cobwebs hanging from tree branches.
A stastical approach applied to 297 orthologous gene clusters in 40 microbial genomes suggests a low rate of interspecies gene transfer. Species relationships can therefore be modeled with a tree structure.
==== Body
Introduction
The role of horizontal gene transfer (HGT) in speciation, adaptation, and evolution of life on earth has been studied intensively [1], and there has been a growing body of evidence of transfers of genes among species [2–4] and transfers from organelles to nuclei [5–7]. Whole genome analyses of different prokaryotes have been thought to indicate rampant HGTs [8,9] and suggest that HGT plays a pivotal role in prokaryotic evolution, producing dynamic and mosaic genomes. The speculation [10] that even genes involved in transcription and translation might have been subject to HGT has also led to the suggestion that HGT should be considered the essence of phylogeny and that HGT might have eroded the organismal genealogical trace. Therefore, life history cannot be properly represented by the traditional treelike form, but rather by a netlike form [4,11–13].
One of the main unresolved issues in the debate is the estimation of HGT frequency [14] and its impact on phylogeny [15]. Commonly used methods for detecting HGT are based on observations of (1) atypical gene sequence composition [16,17]; (2) unexpected rankings for sequence similarity among homologs [18]; and (3) incongruence among phylogenetic trees [e.g., 7]. Studies based on sequence characters suggested HGT frequency at 24% in Thermotoga [2] and a range up to 17% among different prokaryotes [19,20]. Conflicts between the 16S rRNA tree and other gene trees have been frequently reported [e.g., 21]. These findings have led to the ongoing debate about the impact of HGT on phylogeny. Some researchers believe that HGTs are so frequent that a core of nontransferable genes might not exist and that phylogeny in treelike form has little utility [20,22]. Other researchers, however, believe that HGTs constitute only minor interference when inferring phylogeny and propose that methods for inferring HGT have various problems leading to an overestimation of its frequency [1]. For example, a previous study shows that different methods for estimating HGT gave different sets of HGT candidates when applied to the same genome [23]. Meanwhile, it has been proposed that a phylogeny could be sufficiently retrieved via a core of genes that may be resistant to HGT [24,25]. A congruent phylogenetic structure inferred from different genes was proposed as further evidence to buttress this argument [26–29].
As pointed out in Daubin and Ochman [30], there is a difference between assessing genetic transfer among elements with some recognizable homology or orthology to sequences in other genomes and assessing genetic transfer in the entire genome, which may have indeed incorporated significant foreign genetic material, through processes such as selection for pathogenicity [19]. Thus, the key question is whether sequences with recognizable homologs in a significant number of genomes show high levels of HGT and whether HGT's effects are sufficient enough to impede the building of branching phylogenetic history [31]. HGT events lead to incongruent phylogenies for different genetic elements. But at the same time, incongruence in phylogenies can be caused by a list of factors, such as artifacts of phylogenetic reconstruction or other biological sources [32,33]. The inference of HGT from tree comparisons should be done under a proper statistical framework [14]. Furthermore, though HGT events may have occurred with high frequency in genome evolution, perhaps even affecting every gene somewhere in the tree of life, if these events are randomly distributed across the lineages and do not involve more than 50% of the genome at a time, there still exists a backbone tree structure that best fits the majority of the genome. That is, a treelike history will continue to be the most predictive representation of the whole genome and reflect the major mode of genetic information transfer [34], while the HGT events will constitute minor information exchange, much like cobwebs on tree branches. This treelike history of the majority of the genome, which we will call the whole-genome phylogeny, has two utilities. First, barring truly rampant genetic admixture, the tree represents a hypothesis about the major flow of genetic information mediated by the cell replication lineages. Second, such a tree can provide a backbone estimate, from which individual HGT events can be estimated.
Here, we first extend the approach of Lerat et al. [28] and Novichkov et al. [35] with a new method to explicitly test for phylogenetic incongruence due to horizontal transfer versus statistical tree errors, and we subsequently apply it to a larger diversity of genomes. The specific questions we ask in this paper are (1) What is the fundamental structure of the whole-genome (W-G) tree? (2) How do individual gene trees differ from this tree, especially in terms of putative HGT events? (3) How do individual gene trees differ from one another in terms of HGT? (4) What is the rate of HGT events per genome and what kind of genome-specific patterns or gene-specific patterns are evident for HGT events? To answer these questions, we used the Clusters of Orthologous Groups (COG) database of the National Center for Biotechnology Information [36] and extracted the most reliable orthologous clusters. A gene tree was built for each reliable COG and was also integrated to construct a W-G tree. Then, this tree was compared with each gene tree to infer significant HGT events based on our new statistical procedure. We augmented this procedure with a pairwise comparison of gene trees to each other to detect conflicting gene trees. Overall, we find a relatively small proportion of the COG entries with significantly detectable HGT events.
Results
Figure 1 shows our computational flow, which we briefly describe here and in detail in the Materials and Methods section. First, we use the COG database to assemble a set of high-quality orthologous groups for tree inference. Then we use both the combined information in all of the orthologous sequences and the method of Kim and Salisbury [34] to construct the W-G tree that represents the best treelike description for the genealogical relationship of the genomes. Next, we estimate a tree for each orthologous group (which we will refer to as a gene tree) and assess the difference between the tree structure of each gene tree and that of the W-G tree. Next, we augment this comparison with all pairwise comparison of the gene trees. Finally, we evaluate the differences in the tree structure to determine whether these differences can be statistically explained away as general tree errors (e.g., due to noise in the data) or as a set of putative horizontal transfer events. The key point in this procedure is that we explicitly test the alternative hypothesis of HGT rather than merely a rejection of congruence. In the end, for those genes that display statistically significant evidence of horizontal transfer, we estimate the position of the transfer events based on the W-G tree and compute genome-specific and gene-specific rates of transfer.
Figure 1 Flowchart of the HGT Inference Procedure
The National Center for Biotechnology Information COG database is preprocessed for a high-quality COG set. This set is used to construct individual gene trees and the W-G tree, using the median tree algorithm. The gene trees are compared against the W-G tree to detect changes in tree topology that are best explained by a branch transfer. The same comparison is done among all gene trees.
High-Quality Gene Groups and the W-G Tree
The COG database, built by all-versus-all sequence comparisons, covers 43 microorganisms, including complete genomes of bacteria, Archaea, and Saccharomyces cerevisiae, in the initial version, which we used for this study. Our stringent high-quality COG selection procedure described in the Materials and Methods section resulted in the retention of 297 COG entries out of the original 3,852, which cover 40 genomes (Table 1). On average, each high-quality COG covered 16.5 genomes, representing both universally distributed genes and lineage-specific genes. Rather than use any single set of sequences (e.g., rRNA) to approximate the W-G tree, we used the median tree estimator designed by Kim and Salisbury [34], which is a robust estimator that attempts to overcome major genetic distortions such as HGTs. The high-quality COG entries were used to construct the median tree estimate (as shown in Figure 2) with bootstrap values obtained from bootstrap resampling of the input COG entries (the branches with less than 50% bootstrap support were collapsed to improve the reliability of later analysis).
Figure 2 The W-G Tree Based on the Median Tree Algorithm
A subset of high-quality COG entries, which covers at least seven genomes, was used to build the W-G tree (see Materials and Methods). Branches with bootstrap scores less than 50% were collapsed into the polytomous form. Three domains of life are shown as (A) Archaea, (B–J) Bacteria, and (K) Eukaryote. Species are labeled with different colors based on their inferred HGT rates: red, >4%; yellow, 3%–4%; pink, 2%–3%; blue, 1%–2%; green, <1%. Taxonomy labels are (A) Euryarchaea, (B) Proteobacteria, (C) Chlamydiae, (D) Spirochaetes, (E) Thermotogae, (F) Aquificae, (G) Actinobacteria, (H) Deinococcus, (I) Cyanobacteria, (J) Firmicutes, and (K) Fungi.
Table 1 Number of COG Entries That Contain Each of 40 Genomes
In this unrooted W-G tree, the three domains of life are monophyletic with high bootstrap values. Also, the tree strongly supports the monophyly of Chlamydiales, Spirochaetes, low G+C gram-positives, high G+C gram-positives, and α-, β-, γ-, and δ-Proteobacteria. The artifactual attraction of long branches of Archaea and hyperthermophilic bacteria does not appear, with the grouping of Aquifex aeolicus and Thermotoga maritima into the bacterial domain, which is consistent with recent studies [37]. However it should be noted that other authors suggest A. aeolicus should group with Proteobacteria based on shared putative unique indels, protein domain architecture, and membrane structure [38–41], and the grouping remains controversial. Although most of the branches are supported by high bootstrap values, it is worth noting that this tree is partially unresolved, as branches with bootstrap values lower than 50% have been collapsed. Hence, this tree neither informs us on the basal position of the bacterial domain, nor informs us much on the basal branching patterns of archaeal phylogeny, which results in some loss of power for detecting HGT events across these lineages (see also Discussion). Outside of this, two possible artifacts are the basal position of Halobacterium at the archaeal domain, which has been suggested to be affected by a large number of HGT events from bacterial origin [42,43], and the grouping of ɛ-Proteobacteria with Chlamydiales and Spirochaetes, which are commonly seen in literature [27].
Statistical Inference of HGT and Power Test
HGT events for a particular gene or sequence can be detected using phylogenetic methods that compare the estimated gene tree for the candidate sequences against the other gene trees or against some candidate tree that represents the history of the genomes. Mathematical tree distance metrics can be used to measure the discrepancy between two trees. However, two trees may be different because of an HGT event or other reasons, such as noise in the data, compositional bias, hidden gene duplication, gene loss, and so on. Thus, one possible approach is to ask whether the discrepancy between two trees can be more easily explained by simple branch-exchange events (which would be evidence of HGT)—i.e., to explicitly consider the HGT as an alternative hypothesis.
Suppose we have two trees, A and B. If A and B differ by branch-transfer events, they should share a common subtree, wherein the transferred branches have been removed. A bound on the size of such a shared common subtree can be computed using an algorithm called maximum agreement subtree (MAST [44]). Moreover, the difference between A and B can also be measured by the number of branch edges shared by the two trees, computed by a measurement called symmetric difference (SD) metric (also known as Robinson-Foulds metric [45]). Simply, the SD metric computes the number of different splits, regardless of whether or not the difference in the two trees can be explained by a branch switch (and thus putative HGT). Hence, the combination of MAST distance and SD distance between tree A and tree B can be interpreted in terms of putative HGT events (Figure 3). If both MAST and SD distance values are low, then the two trees are not likely to be statistically different. If both MAST and SD values are large, then they may be different, but the difference is not easily explained by an HGT event. The disparity could be due to many other factors (including, of course, HGT). On the other hand, if the two trees differ by a large SD value but are generally similar with a small MAST score, this suggests that the difference can be best explained by putative HGT events. The last case, large MAST distance but low SD distance, cannot occur due to algorithmic reasons.
Figure 3 HGT Inference via Tree Comparison
Raw difference between the SD and the MAST metrics for a given pair of trees tends to increase when HGT is involved in one tree. For example, the raw SD and MAST scores for Gene Tree 1 and the W-G tree are 2 and 2, respectively, while the SD and MAST scores for Gene Tree 2 and the W-G tree are 8 and 2, respectively. This difference between the SD and the MAST scores indicates possible HGT in Gene Tree 2; the (c and d) clade are transferred to the g lineage (dotted arrow). In Gene Tree 1, the (c and d) clade cannot be inferred as transfers because many other factors could have caused the local uncertainty in branching, which should be presented in polytomous form.
Taking this into account, we developed a hypothesis test for HGT, using the difference between the normalized values of the two metrics, which we denote by γ (see Materials and Methods). We computed the significance of an observed γ-value by generating a nonparametric null distribution based on randomly bootstrapped gene trees (see Materials and Methods). In our tree topology-based HGT test, we do not explicitly take branch length into account; however, the bootstrap distribution implicitly allows the incorporation of branch-specific confidence. HGT was inferred when the observed γ was significant with the p-value below the 5% level. The power of this procedure in detecting HGT was tested with a simulation study (detailed in Materials and Methods). These simulation studies applied to each COG showed that on average we were able to detect HGT events at 53.8%, 70.0%, and 77.3%, respectively, for one, two, and three HGT events in a COG tree using the 5% significance value. That is, if the tree contains two HGT events, we can detect the event 70.0% of the time, while guarding against false positive error at the 5% level. We examined the power of our procedure individually for each of the COG datasets; however, we did not observe a significant difference in power between those COG entries where we actually detected HGTs and those where we did not. Therefore, the procedure is not biased toward estimating HGT for one particular kind of tree over another. We also examined the effect of the number of genomes in each COG. Figure 4 shows the results, where the power increases somewhat with larger COG entries, but remains relatively stable. In particular, for those cases where we detected significant HGT events, we do not see a bias toward larger COG entries.
Figure 4 Power of the γ Test in Detecting HGT
Random SPR operations were applied to each COG tree to assess the power of the γ test. The figures show the power values plotted against the taxon numbers in the COG entries for 1, 2, and 3 SPR changes.
HGT Estimation via Comparisons between Each Gene Tree and the W-G Tree
The hypothesis test described above was applied to each of the 297 COG gene trees against the W-G tree. We expected different p-values for the significance level to affect the power of the test, with larger critical p-values tending to more liberally infer HGT events. We investigated the effect of the significance levels on the inferred number of HGT events. The number of COG entries inferred to contain HGT events does not increase dramatically as the cutoff significance value increases (Figure 5). Thus, assuming the standard 5% significance level seemed acceptable to guard against type I error; more liberal values are not expected to significantly change our conclusions about genomic rates of HGT. At the significance level of 0.05, we inferred that 33 out of 297 COG entries (i.e., 11.1%) contain putative HGT events (Table 2). Below, we will call the COG entries with statistically significant HGT events hCOGs. These hCOGs cover a wide range of functional categories as annotated in the COG database [36]. Figure 6 shows the relative frequency of hCOGs within each functional category and aggregated into broader functional categories. We used Fisher's exact test (two-sided) [46] to determine the relationship between the presence of HGT and functional categories. Only one functional category H (coenzyme metabolism) stood out as having a significantly higher (at 0.05 significance level) amount of HGT events. This is in agreement with HGT cases found in literature [4] and supports the speculation that so-called operational genes are more prone to HGT than so-called informational genes [24,47].
Figure 5 The Relationship between Detecting COG Entries with HGT and the p-Values
Dotted curve: the number of COG entries detected to contain HGT at given p-value cutoffs. Straight line: the number of COG entries identified to contain HGT merely by chance, based on given p-value cutoffs. When the cutoff for p-value increases, the number of COG entries that might contain HGT increases, as one would expect. However, the small slope of this curve compared with the line of null hypothesis suggests that the frequency of HGT does not change dramatically, even in a relatively flexible p-value range.
Figure 6 Distribution of Transferred Genes in Different Functional Categories
Functional category abbreviations can be found in Table 2. The percentage of transferred genes in coenzyme metabolism (H) is significantly high, based on Fisher's exact test.
Table 2 List of Transferred Genes
HGT Estimation via Comparisons among Gene Trees
One problem with the above procedure is that the results are sensitive to the particular reference tree, i.e., the W-G tree. To overcome this problem, we next tested for possible HGTs by all pairwise comparisons of 297 COG entries. However, the COG entries do not all share the same taxa, and when the number of shared taxa is too low, we do not have sufficient power to estimate HGTs. Thus, we compared 14,004 pairs of gene trees that contained greater than or equal to six shared taxa. The same hypothesis test for HGT was applied to these pairs of gene trees. With the significance cutoff at 5% level, 1,764 out of the 14,004 pairs were significant under our test, suggesting that 12.6% of the tree pairs contain two trees significantly different from each other in terms of HGT. We then used this fraction to calculate the percentage of hCOGs. In pairwise comparisons, we have the following four cases: (1) neither tree has HGT events; (2) the first tree has HGT events; (3) the second tree has HGT events; and (4) both trees have HGT events. Suppose in our collection, we have x percent of COG entries with detectable HGT events. Then for a given COG, if it is a normal COG, we would expect it to test significantly different in x percent of the comparisons; if it is an hCOG, it should test differently for all of the comparisons. By considering such pairwise tests we can estimate the percentage of the COG entries with detectable HGT events (see Materials and Methods for more details). In our case, we estimate that 13% of COG entries may contain HGT, which is not far from the estimate (11.1%) obtained from W-G tree comparison. The pairwise test may have greater power for discrimination since most of the gene trees are fully resolved compared to our W-G tree.
HGT Frequency in 40 Microbial Genomes
For each of the 33 hCOGs that were identified based on the comparison between each COG tree and the W-G tree, we estimated the positions of putative transfers by using an exhaustive searching procedure (see Materials and Methods for details). This allowed us to compute the genome-specific rate of HGT events among the high-quality COG entries as an estimate of the overall rate of HGT events per genome. Figure 2 shows a colored annotation of the genome-specific rate of HGT laid on top of the W-G phylogeny. Table 3 lists the HGT rate per each genome and the particular COG entries involved in the HGT. The distribution of HGT events along the W-G phylogeny shows no obvious pattern of concentrated events: genomes with high rates of HGT events seem evenly scattered across the phylogeny. As listed in Table 3, the frequency of HGT events ranges from 0% in Chlamydia pneumoniae and six other genomes to 6.7% in Methanobacterium thermoautotrophicum. The rates of HGT in Aeropyrum pernix, Xylella fastidiosa, and some other archaeal organisms, which are notable for their dynamic genome evolution, are relatively high in our result; while the rates for some intensely studied organisms, such as Escherichia coli, are not as high as previously reported [17,20]. Of the top five genomes in our list, all except the M. thermoautotrophicum rank highly for rates of HGT in other surveys [e.g., 17,48,49]. M. thermoautotrophicum, which seems to be typically at the middle of HGT rates in other surveys, stands out in our assay. One possibility is that Dufraigne et al. [48] found M. thermoautotrophicum to have unusually long stretches of putative HGT tracks—and perhaps offering more power by our topology-based test. The mean rate of HGT, 2.0%, among core genes per genome, is considerably lower than those reported in other studies, but the result is consistent with some phylogenetic studies focusing on smaller sets of species [50].
Table 3 Frequency of HGT in 40 Genomes and List of Transferred Genes
Discussion
Our main results show that HGT events can be inferred in only 33 out of the 297 COG entries studied (11.1%) in a comparison against a reference tree and 13% in pairwise comparisons among the tree pairs. The estimated rates of HGT in different genomes are between 0% and 6.7%, with an average of 2.0% among the 40 genomes studied here. There are several factors to consider in this rate estimation. First, as noted in Daubin et al. [51], one of the key questions is the rate of HGT events within those genes that can be orthologously compared to one another reliably (even if they are part of a paralogous family). The use of the COG database and our procedure for retaining only high-quality COG entries mean that our rate computation is limited to such gene sets. Thus, similar to Lerat et al. [28], where very few conflicts among gene trees of widespread single-copy orthologs in γ-Proteobacteria were found, our computed rate of HGT is only for those genes for which reliable orthologous copies can be found in multiple genomes. This might underestimate the HGT rates by ignoring sporadically distributed genes shared by only two or three genomes and those orthologous groups that cannot be reliably assembled via the mutual best-hit approach. On the other hand, for genes from large paralogous families or those only found in a few genomes, reliable assessment is impossible for either HGT or vertical transmission.
Second, we used a specific statistical test where, rather than simply asking whether two gene trees are significantly different from each other, we asked whether the trees are different and can be significantly better explained by horizontal transfer. With this in mind, we conducted a test for a specific alternative hypothesis of HGT rather than the broad rejection of simple tree congruence. A specific test of alternative hypothesis provides additional protection against false rejection of the null hypothesis. Our simulation studies suggest that our test retains reasonable power for detecting HGT events despite this additional precaution. Recently, Novichkov et al. [35] carried out a test for abnormal pairwise divergence patterns similar to our test (but with a stronger assumption of a molecular clock) and found possible HGT in approximately 17%–30% of the COG entries. The fact that we specifically test for positive evidence of HGT and allow more relaxed non-clock-like evolution may explain this discrepancy.
Third, the significance level of the hypothesis test can change the rate estimates. However, within the range of values examined, the estimated numbers of HGT events do not significantly change with increased risk of false rejection. For example, if we increase the significance value to 0.1, then we obtain 39 out of 297 COG entries (13.1%) that may contain HGT events, which is still within the lower range of values reported by others.
Fourth, our statistical test has greater power for phylogenetically distant transfer events compared to proximal transfers. This is because the tree comparison metric SD and MAST differ the most when a tree involves a branch transfer among distant taxa. For nearest-neighbor branch transfers, both methods yield the same value and thus cannot distinguish simple statistical error versus potential HGT event. Hence, if the HGT events frequently involve sister taxa, our estimate of HGT rates will be an underestimate. It is not clear whether HGTs should be more common between close lineages [52]. The mechanism and potential effect of HGT events are different from recombination and hybridization, and therefore it is difficult to assert a lineage distance effect. For example, HGT between distantly related taxa might be argued to be more likely purely due to the increased elapsed time.
Finally, we excluded the high SD and high MAST as cases where HGT events cannot be decided with high confidence. We tested the significance of both high SD and MAST scores using the bootstrap procedures described above. We found just 44 out of 297 cogs (14.8%) that have significantly high SD and MAST values but do not have significantly low γ for both the W-G tree comparison and the pairwise comparison. We are wary of treating such cases as HGT events, but regardless, these cases can be considered to add to an upper bound to HGT estimation. But we believe that such HGT events will be very difficult to detect based on gene genealogies alone. A reliable test would require more densely sampled taxa or other supporting evidence such as sequence compositional characteristics.
We have previously shown by simulation methods that even when there are large-scale HGT events (several events per gene), there remains a recognizable tree that represents the consistent treelike evolution of the majority of the genes and lineages [34]. One way to consider this is to imagine a very large tree, say 10,000 taxa, and some large number of potential “units” of HGT, say 10,000 such elements per genome. Even if each such element had, say, 1,000 actual HGTs across the 10,000 taxa, if we overlay the 10,000 trees on top of one another, all the HGTs will appear as extremely thin connections like cobwebs, and we will see a strong image of a backbone tree. More precisely, consider the relative distance between two taxa as estimated by a set of n genes. Assume that our estimators are perfect; we can obtain exact scaled distance estimates in such a manner that we can estimate the absolute time of separation of the two cell lineages. Let the true time of separation be T
0 and assume that an HGT event along the two lineages yields some variant time estimate, larger if the HGT event brings in a homologous copy from outside the extent of the two lineages, or smaller if the HGT event involves homologous copies from inside the two lineages. Say, for the n genes, k of them experienced horizontal transfer; then we have (n − k) values of T
0, and we need at least as many coincident draws for the HGT time estimates to set some other time estimate to be the modal value—an extremely unlikely event given the possible variant time points in a diverse tree. Thus, while an HGT event can considerably distort the treelike structure of genomic information, there still remains a distinct tree representing the modal information lineage.
Our W-G tree described here is an explicit estimation of this “modal lineage” tree. The estimation of such a modal lineage tree allows us to use explicit tree-based techniques to estimate deviations, i.e., HGT events. When we dissect the signals based on phylogenetic methods with an explicit hypothesis test for HGT, we find that HGT is not as widespread as previously believed [53,54]. Furthermore, the estimated degree of HGT is consistent no matter whether we base it on the modal lineage tree or on pairwise comparisons. The list of HGT candidates is far from being long enough to be called “rampant” for orthologous gene sets, and the overall rates are similar to those found in other studies using phylogenetic methods [28]. We are far from claiming that the reconstruction of the history of life is trivial; however, new developments in orthologous clustering, multiple sequence alignment, tree construction algorithms, and tree-rooting problems may shed more light on the impact of HGT on phylogeny and help us understand the multiple forces of prokaryotic evolution.
Materials and Methods
Input data preparation—Selecting high-quality COG entries
We obtained a set of putative orthologous gene clusters from the COG database (the initial version [36]; ftp://ftp://ftp.ncbi.nih.gov/pub/COG/old/). These data were processed in the following way to increase the reliability of later analysis. (We also excluded three small genomes from the original COG database, as the number of high-quality COG entries covering these genomes was too small.) (1) Best-hit confirmation: all-against-all BLAST searching was redone for all 43 genomes, and every “two-way or one-way best-hit” status for each pair of proteins was tested. Protein members that were not the top hits for any other proteins were removed from the dataset. (2) Removal of large protein families: some of the COG entries are superprotein families, which have gone through extensive gene duplication. They are not easy to use in building reliable sequence alignments and are not suitable for supertree construction. We excluded those large COG entries where the number of sequences exceeds the number of genomes by more than 2.5-fold. (3) Estimation of COG quality by checking BLAST sequence alignments: the quality of each COG was assessed based on the pairwise BLAST e-values and lengths of the significant aligned regions. We obtained 511 COG entries from which all the e-values of pairwise BLAST scores were lower than 10–10, and whose proportion of high-scoring aligned regions to the whole protein sequences was greater than 50%. (4) Building distance matrices: we first generated multiple sequence alignments using CLUSTALW [55], and then used PHYLIP [56] to calculate distance matrices based on the alignments. PHYLIP's command tool “prodist” with default setting of Dayhoff PAM matrix was used to make our calculations. Gap regions in the alignments were dropped because they might not have been aligned properly. (5) Exclusion of paralogs from each COG: in order to have only one representative gene from one genome in each COG, we removed putative paralogous genes. Based on the distance matrix, if the distance between paralogs within one genome was less than the distance to homologs in other genomes (so-called in-paralogs), we randomly chose one of them. If the two paralogs were not closer to each other than to other homologs from other genomes, we filtered out the two paralogs of the same genome from our analysis because they conflicted with each other. After this step, we retained 297 COG entries with at most one sequence per genome. At the same time, these COG entries contain at least six taxa.
Building gene trees and the W-G tree
For each of the 297 COG entries we constructed a gene tree by computing a neighbor-joining tree (PAUP* [57]), using the distance matrix computed as described above. For each gene tree, 1,000 bootstrap replicates were computed by bootstrap sampling from the original sequences and computing a replicate distance matrix. We computed a consensus tree for the bootstrap replicates according to the majority rule; this was used as the gene tree estimate.
We then applied the median tree algorithm [34] to 230 out of the 297 COG entries to build the W-G tree estimate. First, to describe in brief, given a set of distance matrices, the algorithm computes the median of normalized distances as a robust estimate of the true evolutionary distance. It has been shown to be particularly useful for estimating the genome tree when individual genes undergo HGT events. The detailed procedure follows: (1) Data selection: although there were 297 high-quality COG entries, those that covered only a small number of genomes could render the normalization process unstable. Therefore, we used a subset of 230 high-quality COG entries that covered at least seven genomes for the W-G tree estimate. (2) Normalization of distance matrices of COG entries: the median tree algorithm requires us to normalize the distance matrices of COG entries so that we can minimize the difference in evolutionary rates of all the genes in different COG entries. We carried out the normalization to get a scaling factor for each COG in three steps: (i) we selected a single COG that covers all 40 genomes as the reference COG (see Accession Numbers section); (ii) for each COG, we calculated the ratio of pairwise distance for each two genomes to the corresponding pairwise distance in the reference COG; (iii) we used the median of the ratios of the pairwise distances for each COG as the scaling factor for that COG. (3) Building the median tree: for each pair of genomes, the genomic distance was defined as the median of the normalized distances between this pair of genomes for all the 230 normalized COG entries. The median distance of each pair of genomes was retrieved from an average of 45 COG entries, with a minimum of six and a maximum of 147 COG entries (standard deviation = 25). Each COG contributed to an average of 19.7% of the entries in the final median distance matrix, with a minimum of 3.6% and a maximum of 100% (standard deviation = 21.0%). The median distance matrix calculated in this manner was used in conjunction with PAUP* [57] to construct a neighbor-joining tree. One thousand bootstrap replicates of the W-G tree were obtained by bootstrap resampling of the 230 COG entries (the reference COG was guaranteed to be in each resampling), recomputing the median distance matrix, and applying the neighbor-joining method. The majority-rule consensus tree of the bootstrap replicates was computed to estimate the W-G tree (see Figure 2).
HGT inference and power testing
HGT events were tested by computing a statistic γ based on tree topological comparisons. For a pair of trees T and T′, with m and n splits (i.e., branches), respectively, and with x number of taxa, our statistic γ is defined as:
where dS is the SD metric [45], and dM is the MAST metric [44], both of which can be calculated using PAUP* [57]. The terms on the right-hand side are normalized values of the SD and the MAST metrics. The normalization takes into account the effect of the size of the trees on SD and MAST metric. The null distribution described next is also based on the normalized statistics, thus controlling for taxon sampling effects of tree topologies.
The null distribution of γ was obtained by a randomization procedure. For each COG, 2,000 bootstrap trees were generated from the original sequence alignment. We divided them into 1,000 pairs of trees, for each of which the statistic γ was calculated. The distribution of 1,000 γ-values computed in this manner represents the null distribution in which the tree differences are not due to HGT. For each COG, γ was calculated for the gene tree against the W-G tree in the W-G approach or against another gene tree in the pairwise comparison approach, with both trees pruned to the same set of taxa. If the γ-value for a COG was higher than 95% of γ-values for bootstrapped trees, we accepted the alternative hypothesis that HGT is present.
We used random subtree pruning–regrafting (SPR) operations [58] on each COG tree to assess the power of the test, since a gene tree with an HGT is the result of applying a corresponding SPR to the W-G tree. We applied one to three random SPR operations on the COG tree to obtain a changed tree. We repeated this experiment 10,000 times for each combination of COG entries and the number of SPR operations; in this manner we collected the distribution of the γ statistic under the alternative hypothesis of HGTs. This allowed us to obtain the power of our test based on a 0.05 significance level. The average power values for all COG entries are plotted against the number of taxa in the COG entries (see Figure 4). While the power clearly increases with the size of COG entries, the size of the hCOGs does not show a biased distribution, and thus differential power with respect to the numbers of taxa does not seem to be a factor in our results.
Testing the relationship between HGT and functional categories
For each functional category, a 2 × 2 contingency table was built with four elements: (1) number of COG entries with HGT in this category; (2) number of COG entries without HGT in this category; (3) number of COG entries with HGT that are not in this category; (4) number of COG entries without HGT that are not in this category. Two-sided Fisher's exact test was applied to test the association between the HGT and functional category.
Estimation of fraction of COG entries with HGT based on pairwise COG tree comparison
The 297 COG entries may contain overlapped taxa. We compared 14,004 pairs of COG trees that had at least six shared taxa via our γ test. If a given COG contained HGT, we would expect this COG to test positive against all other comparable COG entries. If a given COG is normal, it will only test positive against some unknown fraction P of the hCOGs. Therefore, for each COG we applied our γ test against all other comparable COG entries with a Bonferroni correction for multiple tests. If a COG tested positive against greater than some predetermined P percent of hCOGs, that COG was assigned to the hCOG category. After applying this procedure to the entire set of COG entries, the fraction hCOGs was computed as the value Q. If our procedure is consistent, we should obtain P = Q. Therefore, we iterated through all values of P and repeated our process until P = Q, resulting in an estimated 13% of COG entries ending up in the hCOG category.
Identification of transferred branches in gene trees
The comparison between a gene tree and the W-G tree described above infers presence and absence of HGT events for a given COG. For each COG that tested positive for HGT events, we identified the particular branches of transfer by exhaustive enumeration of possible subtree matches. Since the MAST score gives the number of taxa needed to make the two trees identical, we exhaustively searched for all combinations of branch prunings to find the “troublesome” branches. When there is only one way of pruning branches to make the two trees congruent to each other, those pruned branches are identified as HGT events. However, on a limited number of occasions, there was more than one way of pruning the branches. We treated those branches as equally probable transfers and assigned them a probability weight based on the number of possible prunings. For each genome, the total number of putative HGT events was summed, and the rate of HGT was calculated based on the number of hCOG entries that contained genes from that genome.
Supporting Information
Accession Number
The COG database (http://www.ncbi.nlm.nih.gov/COG/) accession number for the reference COG is COG0541.
This work has been supported in part by National Science Foundation Information Technology Research grant 0334866 and National Institutes of Health/National Institute of General Medical Sciences grant 1-P20-GM-6921–1 to JK.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. JK conceived and designed the experiments. FG and LSW performed the experiments and analyzed the data. FG and JK wrote the paper.
Citation: Ge F, Wang LS, Kim J (2005) The cobweb of life revealed by genome-scale estimates of horizontal gene transfer. PLoS Biol 3(10): e316.
Abbreviations
COGClusters of Orthologous Groups
hCOGClusters of Orthologous Groups database entry containing horizontal gene transfer
HGThorizontal gene transfer
MASTmaximum agreement subtree
SDsymmetric difference
SPRsubtree pruning–regrafting
W-G treewhole-genome tree
==== Refs
References
Brown JR Ancient horizontal gene transfer Nat Rev Genet 2003 4 121 132 12560809
Nelson KE Clayton RA Gill SR Gwinn ML Dodson RJ Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima
Nature 1999 399 323 329 10360571
Koonin EV Makarova KS Aravind L Horizontal gene transfer in prokaryotes: Quantification and classification Annu Rev Microbiol 2001 55 709 742 11544372
Gogarten JP Doolittle WF Lawrence JG Prokaryotic evolution in light of gene transfer Mol Biol Evol 2002 19 2226 2238 12446813
Martin W Stoebe B Goremykin V Hapsmann S Hasegawa M Gene transfer to the nucleus and the evolution of chloroplasts Nature 1998 393 162 165 11560168
Huang CY Ayliffe MA Timmis JN Direct measurement of the transfer rate of chloroplast DNA into the nucleus Nature 2003 422 72 76 12594458
Bergthorsson U Adams KL Thomason B Palmer JD Widespread horizontal transfer of mitochondrial genes in flowering plants Nature 2003 424 197 201 12853958
Raymond J Zhaxybayeva O Gogarten JP Gerdes SY Blankenship RE Whole-genome analysis of photosynthetic prokaryotes Science 2002 298 1616 1620 12446909
Andersson JO Sjogren AM Davis LA Embley TM Roger AJ Phylogenetic analyses of diplomonad genes reveal frequent lateral gene transfers affecting eukaryotes Curr Biol 2003 13 94 104 12546782
Woese CR Olsen GJ Ibba M Soll D Aminoacyl-tRNA synthetases, the genetic code, and the evolutionary process Microbiol Mol Biol Rev 2000 64 202 236 10704480
Doolittle WF Phylogenetic classification and the universal tree Science 1999 284 2124 2129 10381871
Nesbo CL Boucher Y Doolittle WF Defining the core of nontransferable prokaryotic genes: The euryarchaeal core J Mol Evol 2001 53 340 350 11675594
Woese CR On the evolution of cells Proc Natl Acad Sci U S A 2002 99 8742 8747 12077305
Kurland CG Canback B Berg OG Horizontal gene transfer: A critical view Proc Natl Acad Sci U S A 2003 100 9658 9662 12902542
Philippe H Douady CJ Horizontal gene transfer and phylogenetics Curr Opin Microbiol 2003 6 498 505 14572543
Lawrence JG Ochman H Molecular archaeology of the Escherichia coli genome Proc Natl Acad Sci U S A 1998 95 9413 9417 9689094
Nakamura Y Itoh T Matsuda H Gojobori T Biased biological functions of horizontally transferred genes in prokaryotic genomes Nat Genet 2004 36 760 766 15208628
Olendzenski L Liu L Zhaxybayeva O Murphey R Shin DG Horizontal transfer of archaeal genes into the deinococcaceae: Detection by molecular and computer-based approaches J Mol Evol 2000 51 587 599 11116332
Ochman H Lawrence JG Groisman EA Lateral gene transfer and the nature of bacterial innovation Nature 2000 405 299 304 10830951
Garcia-Vallve S Romeu A Palau J Horizontal gene transfer in bacterial and archaeal complete genomes Genome Res 2000 10 1719 1725 11076857
Klenk HP Meier TD Durovic P Schwass V Lottspeich F RNA polymerase of Aquifex pyrophilus Implications for the evolution of the bacterial rpoBC operon and extremely thermophilic bacteria J Mol Evol 1999 48 528 541 10198119
Bapteste E Boucher Y Leigh J Doolittle WF Phylogenetic reconstruction and lateral gene transfer Trends Microbiol 2004 12 406 411 15337161
Ragan MA On surrogate methods for detecting lateral gene transfer FEMS Microbiol Lett 2001 201 187 191 11470360
Jain R Rivera MC Lake JA Horizontal gene transfer among genomes: The complexity hypothesis Proc Natl Acad Sci U S A 1999 96 3801 3806 10097118
Eisen JA Horizontal gene transfer among microbial genomes: New insights from complete genome analysis Curr Opin Genet Dev 2000 10 606 611 11088009
Matte-Tailliez O Brochier C Forterre P Philippe H Archaeal phylogeny based on ribosomal proteins Mol Biol Evol 2002 19 631 639 11961097
Daubin V Gouy M Perriere G A phylogenomic approach to bacterial phylogeny: Evidence of a core of genes sharing a common history Genome Res 2002 12 1080 1090 12097345
Lerat E Daubin V Moran NA >From gene trees to organismal phylogeny in prokaryotes: The case of the γ-Proteobacteria PLoS Biol 2003 1 e19 10.1371/journal.pbio.0000019 12975657
Dutilh BE Huynen MA Bruno WJ Snel B The consistent phylogenetic signal in genome trees revealed by reducing the impact of noise J Mol Evol 2004 58 527 539 15170256
Daubin V Ochman H Quartet mapping and the extent of lateral transfer in bacterial genomes Mol Biol Evol 2004 21 86 89 12949130
Kyrpides NC Olsen GJ Archaeal and bacterial hyperthermophiles: Horizontal gene exchange or common ancestry? Trends Genet 1999 15 298 299 10431189
Maddison WP Gene trees in species trees Syst Biol 1997 46 523 526
Slowinski JB Page RD How should species phylogenies be inferred from sequence data? Syst Biol 1999 48 814 825 12066300
Kim J Salisbury BA A tree obscured by vines: Horizontal gene transfer and the median tree method of estimating species phylogeny Pac Symp Biocomput 2001 6 571 582
Novichkov PS Omelchenko MV Gelfand MS Mironov AA Wolf YI Genome-wide molecular clock and horizontal gene transfer in bacterial evolution J Bacteriol 2004 186 6575 6585 15375139
Tatusov RL Natale DA Garkavtsev IV Tatusova TA Shankavaram UT The COG database: New developments in phylogenetic classification of proteins from complete genomes Nucleic Acids Res 2001 29 22 28 11125040
Brochier C Philippe H Phylogeny: A non-hyperthermophilic ancestor for bacteria Nature 2002 417 244 12015592
Gupta RS Griffiths E Critical issues in bacterial phylogeny Theor Popul Biol 2002 61 423 434 12167362
Griffiths E Gupta RS Signature sequences in diverse proteins provide evidence for the late divergence of the order Aquificales
Int Microbiol 2004 7 41 52 15179606
Iyer LM Koonin EV Aravind L Evolution of bacterial RNA polymerase: Implications for large-scale bacterial phylogeny, domain accretion, and horizontal gene transfer Gene 2004 335 73 88 15194191
Cavalier-Smith T The neomuran origin of archaebacteria, the negibacterial root of the universal tree and bacterial megaclassification Int J Syst Evol Microbiol 2002 52 7 76 11837318
Brochier C Forterre P Gribaldo S Archaeal phylogeny based on proteins of the transcription and translation machineries: Tackling the Methanopyrus kandleri paradox Genome Biol 2004 5 R17 15003120
Kennedy SP Ng WV Salzberg SL Hood L DasSarma S Understanding the adaptation of Halobacterium species NRC-1 to its extreme environment through computational analysis of its genome sequence Genome Res 2001 11 1641 1650 11591641
Goddard W Kubicka E Kubicki G McMorris FR The agreement metric for labeled binary trees Math Biosci 1994 123 215 226 7827420
Robinson DF Foulds LR Comparison of phylogenetic trees Math Biosci 1981 53 131 147
Conover WJ Practical nonparametric statistics, 3rd ed 1999 New York Wiley 584
Harris JK Kelley ST Spiegelman GB Pace NR The genetic core of the universal ancestor Genome Res 2003 13 407 412 12618371
Dufraigne C Fertil B Lespinats S Giron A Deschavanne P Detection and characterization of horizontal transfers in prokaryotes using genomic signature 2005 Nucleic Acids Res 33(1) e6 Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=15653627 . Accessed 19 July 2005
Medrano-Soto A Moreno-Hagelsieb G Vinuesa P Christen JA Collado-Vides J Successful lateral transfer requires codon usage compatibility between foreign genes and recipient genomes Mol Biol Evol 2004 21 1884 1894 15240837
Ortutay C Gaspari Z Toth G Jager E Vida G Speciation in Chlamydia Genomewide phylogenetic analyses identified a reliable set of acquired genes J Mol Evol 2003 57 672 680 14745536
Daubin V Moran NA Ochman H Phylogenetics and the cohesion of bacterial genomes Science 2003 301 829 832 12907801
Lawrence JG Hendrickson H Lateral gene transfer: When will adolescence end? Mol Microbiol 2003 50 739 749 14617137
Salzberg SL White O Peterson J Eisen JA Microbial genes in the human genome: Lateral transfer or gene loss? Science 2001 292 1903 1906 11358996
Snel B Bork P Huynen MA Genomes in flux: The evolution of archaeal and proteobacterial gene content Genome Res 2002 12 17 25 11779827
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
Felsenstein J PHYLIP (Phylogeny Inference Package), version 3.5c [computer program] 1993 Seattle University of Washington Department of Genetics Available: http://evolution.genetics.washington.edu/phylip.html . Accessed 19 July 2005
Swofford DL PAUP*: Phylogenetic analysis using parsimony (*and other methods), version 4 [computer program] 2002 Sunderland (Massachusetts) Sinauer
Semple C Steel M Phylogenetics 2003 New York Oxford University Press 239
|
16122348
|
PMC1233574
|
CC BY
|
2021-01-05 08:21:26
|
no
|
PLoS Biol. 2005 Oct 30; 3(10):e316
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030316
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030345SynopsisCell BiologyDevelopmentSystems BiologyParasitologyNematodesCaenorhabditisThe Molting Worm Sheds Its Genetic Secrets Synopsis10 2005 30 8 2005 30 8 2005 3 10 e345Copyright: © 2005 Public Library of Science.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.
Functional Genomic Analysis of C. elegans Molting
==== Body
Anyone who winces remembering the challenges of childhood might consider the fate of the developing worm. Four times in its life, a growing nematode worm flips on its side and writhes around to shed its exoskeleton, or cuticle. During each molt, a worm casts aside its cuticle and synthesizes a new protective shell, its primary defense against a harsh environment. Though it's clear that a complex array of signaling proteins and enzymes are required to engineer these rites of passage, only a few genes have been implicated in the process.
Scientists have found critical molting genes in the fruitfly, but most of these are not present in the worm. It may be that the worm has unique molting genes, because its cuticle is more elastic than the hardened casing of an insect. If scientists find molting genes that exist only in the worm, they can begin to unravel the mechanisms that govern these critical phases of a worm's life. They can also develop treatments that target the worm's parasitic cousins—which wreak havoc on humans, livestock, and plants—without producing harmful side effects. In a new study, Alison Frand, Sascha Russel, and Gary Ruvkun searched the entire genome of the worm Caenorhabditis elegans for molting genes and identified 159 candidates, using a technique called RNA interference.
In RNA interference, researchers use double-stranded RNA (dsRNA) to block the expression of, or silence, a specific gene by destroying the gene's messenger RNA transcript before it can be translated into protein. These dsRNAs can be expressed in bacteria—a staple food for lab worms—which then multiply into large colonies expressing the same dsRNA. Using a pre-existing “library” of bacterial clones that each express a particular dsRNA, the authors fed groups of larvae one bacterial clone at a time—until thousands of larvae had eaten bacteria with dsRNA designed to silence nearly every one of the worm's 19,427 genes.
After the larvae ate the bacterial clones, the authors screened them for molting defects—which is how they identified the 159 genes. Molting defects mostly left larvae trapped in their old cuticle; those that managed to escape often failed again during the next molt. The majority of candidate molting genes appear to play a role in all four molts, the authors argue, since their inactivation foils molting at several stages. And, significantly, the majority of genes—many of which are found only in worms—exist in worm parasites that infect humans, animals, and plants.
Among the genes identified, the authors found several transcription factors (proteins that activate genes), indicating that molting requires “extensive changes in gene expression.” Other genes are associated with signaling proteins that likely coordinate the activity of different cell types during molting, and many genes code for proteins that are required for protein synthesis—likely to build the new cuticle. Still other genes may help remodel the cuticle.
As this Caenorhabditis elegans larva molts between developmental stages, green fluorescent protein allows researchers to trace the expression of one of its molting genes (mlt-11). Defects in this gene trap larvae in their cuticle
To monitor the expression of some of these genes and infer their function, the authors tagged a subset of genes—representing many of the functional categories found in the screen—with green fluorescent protein. Since green fluorescent protein glows when a gene is activated, the researchers can see where and when genes are expressed. Fluorescence levels were high just before each molt and dropped off soon after. All of these genes were expressed in the epithelial cells that secrete new cuticle. These results provide strong evidence for the genes' role in molting, since they were expressed both at the right time and the right place. These experiments also allowed Frand et al. to propose a model describing the timing and order of gene expression during molting.
With all the genes identified in this screen, researchers can now start to piece together the overlapping pathways that guide the worm through its formative years. And with the discovery of worm-specific genes, it's likely that more effective treatments await patients with elephantiasis, African river blindness, and other diseases caused by pathogenic nematodes. —Liza Gross
|
0
|
PMC1233575
|
CC BY
|
2021-01-05 08:21:26
|
no
|
PLoS Biol. 2005 Oct 30; 3(10):e345
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030345
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030346SynopsisBioinformatics/Computational BiologyCell BiologyGenetics/Genomics/Gene TherapyMolecular Biology/Structural BiologySystems BiologySaccharomycesA Global View of DNA-Packing Proteins Cracks the Histone Code Synopsis10 2005 30 8 2005 30 8 2005 3 10 e346Copyright: © 2005 Public Library of Science.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.
Single-Nucleosome Mapping of Histone Modifications in S. cerevisiae
==== Body
In one of biology's most impressive engineering feats, specialized proteins package some six-and-a-half feet of human DNA into a nucleus that averages just 5 microns (0.0001969 inches) in diameter. In the first of a series of supercondensing steps, DNA winds around proteins called histones, which together form a complex called the nucleosome. Histones package DNA into repetitive coils, which not only provide genomic structure but also help regulate gene expression. These tasks are mediated in part by chemical modifications to histone proteins—most commonly to histone “tails,” long, unstructured chains of amino acids that protrude from nucleosomes. Different chemical modifications are associated with different functional effects. Acetylation, which adds an acetyl group to an amino acid on the histone tail, has been linked to both gene activation and silencing, depending on which amino acid is modified. Methylation (addition of a methyl group to the histone tail) has also been linked to gene activation and repression, although the chemical effects of methylation differ dramatically from those of acetylation.
Even in yeast, amino acid modifications in the histone tails can number in the tens and twenties. Given the number of possible permutations of modification types and amino acids, the question arose, might different combinations of histone modifications produce discrete outcomes? The notion that a sequence or combination of specific modifications on histone tails acts as a signal to other proteins and produces distinct biological effects was advanced as the “histone code” hypothesis in 2000.
Histones can undergo many potential modifications, and it has been hypothesized that these can occur in many different combinatorial histone modification patterns (A). In this study, researchers found that only a few modification patterns occur in yeast, with many of the modifications co-occurring in groups (B)
Progress in deciphering the vocabulary, mechanics, and function of the histone code has been hindered by the coarse resolution of available tools. Nucleosomes typically cover about 146 base pairs, but existing technology could only average over 500 to 1,000 base pairs at a time—confounding the effects of single nucleosomes. In a new study, Oliver Rando and colleagues take advantage of the high resolution afforded by their custom-made microarray, which has a resolution of 20 base pairs. Working with the budding yeast Saccharomyces cerevisiae, the scientists examined 12 different histone modifications in individual nucleosomes and found only a small number of distinct combinations with “few discrete histone modification patterns.” The concurrent modifications fall into two categories: one set targets a transcriptional start site but is the same no matter what the level of transcription, while the other occurs throughout gene coding regions and is linked to transcription. Importantly, the only modifications that appear to correlate with transcription occur over transcribed regions, as though they were the consequence, rather than the cause, of transcription.
Why might histone tails exhibit so many modifications if they form only two independent categories? It's possible that histone-modifying enzymes may work best in groups and so the marks that recruit them—acetyl and methyl groups—also come in groups. Another possible explanation relates to how histone modifications signal transcription enzymes that a particular gene requires more or less transcription. When the positively charged amino acid lysine acquires an acetyl group, it loses its charge, and charge–charge interactions play a major role in many interactions between proteins and other molecules. Multiple lysine acetylations on the histone tail may thereby aid certain chemical reactions necessary for transcription in a continuous way; having multiple levels of acetylation, for example, may allow the cell to “tune” protein–protein interactions, and thus gene expression, up and down, rather than simply turn it on or off.
Rando and colleagues propose that the histone modifications associated with transcription may facilitate rather than trigger gene expression, perhaps by clearing a path for the transcription machinery or attracting proteins needed for the job. The authors are careful to point out, however, that histone modifications may also play some role in initiating gene expression, but that any transcription pattern would likely be obscured, or “erased,” as transcription occurs. While future studies will help determine which role proves more common, these results suggest that histone modifications are facilitators rather than activators and that the histone code is more a transcription footprint than a starting signal. —Liza Gross
|
0
|
PMC1233576
|
CC BY
|
2021-01-05 08:21:26
|
no
|
PLoS Biol. 2005 Oct 30; 3(10):e346
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030346
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030347SynopsisBioinformatics/Computational BiologyEvolutionGenetics/Genomics/Gene TherapyEubacteriaArchaeaComparing Gene Trees and Genome Trees: A Cobweb of Life? Synopsis10 2005 30 8 2005 30 8 2005 3 10 e347Copyright: © 2005 Public Library of Science.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 Cobweb of Life Revealed by Genome-Scale Estimates of Horizontal Gene Transfer
==== Body
The tree of life has long served as a useful tool for describing the history and relationships of organisms over evolutionary time. One species is represented as a branching point, or node, on the tree, and the branches represent paths of descent from a parental node. The tree diagram carries an implicit assumption that genes are transferred vertically, from parent to child, and that all the genes in a new species come from the ancestral species. In theory, one should be able to trace the origin of each gene in a species back to its ancestor. In practice, however, the ancestral gene is rarely available, so researchers look for the gene in a closely related species. (These similar genes, which diverge slightly after a speciation event, are called orthologs.)
Horizontal gene transfers—gene exchange between non-related organisms—appear commonplace among bacteria, but contribute just small bits of genetic information, leaving the traditional tree of life intact
But as the tools of genome analysis became more refined, searches for similar genes sometimes turned up sequences that belonged to a species on a different branch of the evolutionary tree. Clearly, vertical gene transfer was not the only mechanism of genetic transmission. Organisms, it turns out, can acquire genes from non-ancestral species through a mechanism called horizontal gene transfer (HGT)—think of it as acquiring genes from your neighbor instead of your parents. Such genetic exchanges, most common among bacteria and other microbes, are not represented in the tree of life—no single branch connects the two unrelated species. Initial studies suggested that HGT events were extremely common, prompting some to say it was time to replace the tree with a netlike diagram. Other studies have since suggested that methods used to calculate HGT overestimated its frequency: researchers detect HGT events by finding inconsistencies between gene trees and organism, or whole-genome, trees, but statistical errors can artificially increase the number of HGT events.
In a new study, Fan Ge, Li-San Wang, and Junhyong Kim estimate the frequency of HGT events by using a novel statistical method to compare the gene trees and whole-genome trees of microbes. Their method solves the statistical problem by directly testing for discrepancies between trees that arise from statistical error versus true HGT events. Analyzing over 40 microbial genomes, Kim and colleagues estimate that HGT infiltrates just 2% of the average microbial genome. Even when relatively common, the authors conclude, HGT events do not disrupt the integrity of the tree of life, contributing just small bits of genetic material, “much like cobwebs on tree branches.”
To construct both gene trees and a whole-genome tree for the microbes, the authors selected core sets of orthologous gene groups from the NIH database of clusters of orthologous genes. (Clusters are derived by comparing protein sequences encoded in complete genomes, which represent major lineages on the evolutionary tree. Each cluster corresponds to an ancient, conserved protein domain.) Kim and colleagues created gene trees for each cluster of orthologous genes they selected, then created whole-genome trees from the gene trees and compared each gene tree to the whole-genome tree, using their new method. HGT events were detected when two species appeared close together on a gene tree but far apart on the whole-genome tree. Overall, just over 11% of the orthologous gene clusters showed statistically significant HGT events, with HGTs accounting for about 2% on average of each of the 40 microbial genomes.
Altogether, these results suggest that HGT is not as common as once thought. And even when large-scale HGT events do occur—which Kim simulated in a previous study—they do not obscure the evolutionary path of most genes and lineages. If you imagine a tree with 10,000 taxa, the authors explain, and 1,000 HGTs per genome across all the taxa, the HGTs would form “extremely thin connections like cobwebs,” leaving the backbone of the tree intact. Infrequent though it may be, HGT likely has some impact on the evolutionary history of life—impacts that advances in genome analysis technology may help uncover. —Liza Gross
|
0
|
PMC1233577
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Oct 30; 3(10):e347
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030347
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1617612210.1371/journal.pbio.0030351Research ArticleCell BiologyMolecular Biology/Structural BiologyNeuroscienceDanio (Zebrafish)Temperature Regulates Transcription in the Zebrafish Circadian Clock Temperature Regulates Zebrafish ClockLahiri Kajori
1
Vallone Daniela
1
Gondi Srinivas Babu
1
Santoriello Cristina
1
Dickmeis Thomas
1
Foulkes Nicholas S [email protected]
1
1Max-Planck Institut für Entwicklungsbiologie, Tübingen, GermanyStemple Derek Academic EditorWellcome Trust Sanger InstituteUnited Kingdom11 2005 27 9 2005 27 9 2005 3 11 e35124 3 2005 12 8 2005 Copyright: © 2005 Lahiri 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.
Temperature Regulates the Zebrafish Clock
It has been well-documented that temperature influences key aspects of the circadian clock. Temperature cycles entrain the clock, while the period length of the circadian cycle is adjusted so that it remains relatively constant over a wide range of temperatures (temperature compensation). In vertebrates, the molecular basis of these properties is poorly understood. Here, using the zebrafish as an ectothermic model, we demonstrate first that in the absence of light, exposure of embryos and primary cell lines to temperature cycles entrains circadian rhythms of clock gene expression. Temperature steps drive changes in the basal expression of certain clock genes in a gene-specific manner, a mechanism potentially contributing to entrainment. In the case of the per4 gene, while E-box promoter elements mediate circadian clock regulation, they do not direct the temperature-driven changes in transcription. Second, by studying E-box-regulated transcription as a reporter of the core clock mechanism, we reveal that the zebrafish clock is temperature-compensated. In addition, temperature strongly influences the amplitude of circadian transcriptional rhythms during and following entrainment by light–dark cycles, a property that could confer temperature compensation. Finally, we show temperature-dependent changes in the expression levels, phosphorylation, and function of the clock protein, CLK. This suggests a mechanism that could account for changes in the amplitude of the E-box-directed rhythm. Together, our results imply that several key transcriptional regulatory elements at the core of the zebrafish clock respond to temperature.
Reveals the molecular basis by which temperature cycles entrain circadian rhythms of clock gene expression in zebrafish
==== Body
Introduction
The circadian clock plays a central role in adapting the physiology of plants and animals to anticipate day–night environmental changes. Amongst the most conserved properties of the clock is the ability of daily temperature cycles and acute temperature changes to set its phase [1]. In addition, the period length of the clock rhythm remains relatively constant over a wide range of temperatures [1,2]. The mechanism underlying this “temperature compensation” corrects for the natural tendency of the rate of biochemical reactions to change with temperature. Outside of the range of temperature compensation, the clock stops running and arrests at a certain phase [1,3,4]. The physiological range for rhythmicity typically lies well within the temperature range permissive for growth.
In ectotherms, where core body temperature is strongly influenced by the environment, these properties have clear importance to provide a mechanism for daily entrainment of the pacemaker, as well as to ensure that seasonal variations in temperature do not lead to deleterious changes in the speed of the clock cycle [1,5,6]. Although there is homeostatic control of core body temperature in endotherms, recent tissue and cell culture studies have confirmed that their clocks are also temperature-compensated and can be phase-shifted by acute temperature changes [7–9]. In addition, daily rhythms of body temperature have been directly implicated in the maintenance of peripheral clock function [10,11]. Thus, regulation by temperature appears to be a highly conserved property of the circadian timing system.
Molecular studies in a wide range of model organisms have revealed that many clock genes are components of transcription translation feedback loops [12]. For example, in vertebrates, the basic helix-loop-helix Per-Arnt-Sim domain transcription factors, Clock (CLK) and Brain and muscle Arnt-like protein (BMAL), bind as heterodimers to E-box enhancers and activate the expression of other clock genes that encode transcriptional repressors, the Period (Per) and Cryptochrome (Cry) proteins. These repressors interact with CLK-BMAL and interfere with transcriptional activation, thereby reducing expression of their own genes and so closing the feedback loop [13]. Our limited understanding of the molecular basis of temperature responses of the clock has come from studies of the expression of the period and timeless genes in Drosophila, and the frequency gene in Neurospora. Here, temperature-dependent alterations in transcription [14], mRNA processing [15–17], translation [3,14], protein stability [14,18], and protein–protein interactions [19] have been documented. These results would appear to suggest that unlike the situation for light, multiple regulatory components of the circadian clock mechanism perceive and respond to temperature changes.
Remarkably, little is known about the genetic and molecular basis of the response of the vertebrate clock to temperature. Zebrafish represent an attractive model to explore this issue. Adult fish, larvae, and embryos remain viable over a wide range of core body temperatures (10 °C) that can be regulated accurately by simply controlling the temperature of the water [20]. Unlike other ectothermic vertebrates that have been used in the past to study temperature responses of the clock, zebrafish offer a powerful combination of molecular, genetic, and cell culture tools. Furthermore, the direct entrainment of peripheral clocks in zebrafish by light provides a model to investigate how two zeitgebers integrate to regulate clock function at the cellular level [21,22]. In a recent study of luciferase reporter transgenic zebrafish, Kaneko and colleagues reported that the ambient temperature affects the levels and amplitude of cycling per3 expression, indicating that temperature can influence clock gene expression in this species [23].
Here we have investigated the effects of temperature on the zebrafish circadian clock in more detail. Specifically, we demonstrate that temperature cycles of as little as 2 °C are sufficient to entrain the zebrafish clock. Temperature steps drive significant changes in the expression levels of several clock genes in a gene-specific manner. In the case of the per4 gene, while E-box enhancer elements direct circadian expression rhythms, they do not mediate the decreases and increases in expression that immediately follow acute temperature changes. Also, we have studied how temperature affects light–dark (LD) cycle-entrained rhythmic transcription that is directed by the core clock mechanism via E-box elements. We confirm that the zebrafish circadian clock is temperature-compensated. Furthermore, temperature influences the amplitude of rhythmic transcription, a property that could contribute to temperature compensation. The CLK protein expression levels, phosphorylation, and binding to E-box elements, as well as CLK transactivation efficiency, are all temperature-regulated. These temperature-regulated changes in CLK protein function may contribute to the observed changes in rhythm amplitude.
Results
Temperature Cycles Influence Clock Gene Expression in Zebrafish Embryos and Cells
It has been well-documented that exposure to LD cycles is essential for the establishment of circadian rhythms in zebrafish larvae [24,25]. We asked whether, in the absence of light, temperature cycles could also entrain the zebrafish circadian clock. Zebrafish larvae were raised from the late blastula stage (4 h post-fertilization) in constant darkness (DD) for 6 d, while being exposed to a 24-h, 4 °C temperature cycle (between 23.5 °C and 27.5 °C) and were then sacrificed at three hourly time points. Sibling larvae were raised at a constant temperature (25.3 °C) under LD cycle conditions and sacrificed in parallel. We then assayed the daily RNA expression profile of a representative group of core clock gene homologs (per2 & 4, cry2a & 3, and clock1) [26–29] by RNAse protection assay (RPA). In the case of per4, cry3, clock1, and cry2a, rhythmic expression was detected under both temperature and LD cycle conditions (Figure 1A–1D). Expression of per4 and cry3 peaked at, and just before, the dark–light transition, respectively (Figure 1A) and in the middle of the cold period (Figure 1B). Expression of clock1 and cry2a peaked in the first half of the dark period and at the light–dark transition, respectively (Figure 1C) and at the high–low temperature transition and the middle of the warm period, respectively (Figure 1D). Thus the phase relationship between the expression rhythms of all four genes is preserved under LD and temperature cycles, with the high temperature phase roughly corresponding to the light period. Remarkably, larvae raised in a 2 °C daily temperature cycle gave comparable results (Figure S1A). Thus, exposure of zebrafish to even shallow temperature cycles during early development is sufficient to establish rhythmic clock gene expression. In contrast, while per2 rhythmic expression is encountered in the LD cycle larvae, low, non-oscillating levels are detected in the temperature cycle conditions (Figure 1E), in agreement with previous reports showing that per2 expression is light-driven [30,31]. This reveals a differential response of clock gene expression during exposure to LD and temperature cycles.
Figure 1 Rhythmic Clock Gene Expression under LD and Temperature Cycles
Graphical summary of RPA assays are described:
(A) Per4 (solid line) and cry3 mRNA expression (dashed line) in zebrafish larvae raised for 6 d either in a light (12 h) or dark (12 h) cycle at a constant temperature (25.3 °C).
(B) Per4 (solid line) and cry3 mRNA expression (dashed line) in zebrafish larvae raised for 6 d in DD, under a temperature cycle of 4 °C (23.5 °C/11 h, 27.5 °C/11 h, plus 1 h for each heating and cooling phase). RNA samples were harvested during the seventh day (ZT0 is defined as the beginning of the heating and light periods).
(C and D) Equivalent analysis of clock1 (solid line) and cry2a (dashed line) expression in (C) LD, and (D) temperature cycle larvae.
(E) Per2 expression was assayed in LD (dashed line) or temperature cycle (ΔT) larvae (solid line). By linear regression analysis, the slope of the ΔT trace has no significant deviation from zero (R2 = 0.033 and p = 0.66, F-test). The LD cycle curve fits to a 6th-order polynomial regression model (R2 = 0.96 and Runs test for deviation from model p = 0.99).
In each case, zeitgeber time is plotted on the x-axis while the relative expression levels (percentage) are plotted on the y-axis. β-actin levels were used to standardize the results. The highest band intensity in each experiment was arbitrarily defined as 100%, and then all other values were expressed as a percentage of this value. All experiments were performed in triplicate, and error bars denote the standard deviation.
We have previously demonstrated that zebrafish primary cell lines are powerful in vitro tools to explore the molecular basis of how light entrains the clock in zebrafish peripheral tissues [22,27]. Could these cells also be useful for studying the effects of temperature on the clock? We tested whether exposure to a 4 °C temperature cycle in DD induces rhythmic clock gene expression in the zebrafish PAC-2 cell line. Consistent with our larvae results, we observed an equivalent pattern of clock gene expression in the PAC-2 cells (unpublished data and Figure S1B). Control cells that remained in constant temperature conditions during the entire experiment failed to show expression rhythms. These results indicate that the clock mechanism in zebrafish primary cell lines is able to directly perceive and respond to temperature as well as lighting changes.
Temperature Cycles Entrain the Zebrafish Circadian Clock
An important issue is whether temperature cycles drive changes in clock gene expression or if they actually entrain the zebrafish circadian oscillator. We addressed this question in the following ways by examining expression of the per4 gene in PAC-2 cells. One prediction for entrainment of the oscillator is that circadian rhythms of clock gene expression established by temperature cycles should persist for several cycles following transfer to a constant temperature. We confirmed that a circadian rhythm of endogenous per4 expression could be measured between 24 h and 72 h after transfer of cells from a 4 °C temperature cycle to constant temperature (Figure 2A). Control cells held at a constant temperature during the entire experiment failed to show cycling gene expression (unpublished data and Figure S1B). We also tested how the phase of clock gene expression is affected by changing the thermoperiod of the temperature cycle (i.e., the relative length of the warm and cold phases of the 24-h cycle). The prediction for entrainment is that the phase of the clock gene rhythm will change relative to the phase of the temperature cycle as a function of the thermoperiod, i.e., a phase angle change [32]. In a temperature-driven response, the phase relationship would be predicted to remain constant; as the cells are warmed or cooled, a change in gene expression should occur immediately, regardless of the thermoperiod. [32]. We exposed PAC-2 cells to 24-h temperature cycles, with either short (8 h) or long (16 h) warm periods and then assayed per4 expression. The daily expression peak shifts from 9 h after the warm–cold transition in 8:16 (8 h warm:16 h cold) cycles to 5 h after the same transition in 16:8 cycles, and in this way remains consistently locked to the middle of the cold period (Figure 2B). Thus, also this result points to the clock being entrained by temperature cycles. Interestingly, the start of the increase in per4 expression consistently coincides with the warm–cold transition, with differences in the waveform leading to changes in the timing of the subsequent peak. This suggests that temperature steps can also drive changes in clock gene expression.
Figure 2 Temperature Cycles Entrain the Zebrafish Circadian Clock
(A) Quantification of per4 expression levels as measured by RPA, in WT PAC-2 cells initially exposed for 6 d to a 4 °C temperature cycle and then transferred to a constant temperature (25.5 °C) for 72 h. RNA extracts were prepared from 24 h to 72 h following transfer to the constant temperature at four hourly intervals. RPA band intensities were quantified and adjusted as described in Figure 1.
(B) RPA results quantifying per4 expression levels in WT PAC-2 cells exposed for 5 d to warm:cold temperature cycles (as indicated below the x-axis), including either a 8 h:16 h (blue trace), or a 16 h:8 h (red trace) cycle. The blue and red dashed lines and arrowheads indicate the delay between the warm–cold transition and the peak of per4 rhythmic expression in each temperature cycle. All experiments were performed in triplicate, and error bars denote the standard deviation.
Temperature Shifts Directly Drive Clock Gene Transcription
Previous results have demonstrated that temperature steps can elicit acute changes in the expression levels of clock genes in non-vertebrates [14]. Do zebrafish also show this property? We raised larvae at a constant temperature in DD, then abruptly shifted the temperature up or down by 8 °C (21 °C → 29 °C, or 29 °C → 21 °C, respectively), and the resulting expression of a range of clock genes was assayed during the following 6 h (Figure 3A and 3B). Given the strong influence of temperature on the rate of early zebrafish development [20], the larvae at 21 °C were raised for 2 d more than those at 29 °C prior to the temperature step in order to ensure that both sets were at a comparable developmental stage. Expression of per4 and cry3 was strongly down-regulated following the temperature increase (3.3 ± 0.4-fold and 1.7 ± 0.2-fold, respectively, at 4 h following the temperature step up), and up-regulated following the temperature decrease (4.7 ± 0.8-fold and 3.6 ± 0.6-fold, respectively, at 4 h following the temperature step down), while the opposite response was observed for cry2a (1.8 ± 0.2-fold increase and 2.0 ± 0.3-fold decrease at 4 h following the temperature steps). The temperature steps did not affect the expression of clock1, per2, or β-actin. Furthermore, no significant changes in expression for any of the genes were detected in control samples that remained at a constant temperature (21 °C or 29 °C) in DD (Figure 3A and 3B). These results reveal that changes in zebrafish clock genes expression do occur following temperature steps, and that this constitutes a gene-specific response.
Figure 3 Temperature Steps Regulate Clock Gene Expression Levels
(A) Larvae were raised in DD at 21 °C for 7 d and then shifted to 29 °C and harvested at the indicated times relative to the temperature shift (h). Controls remained at 21 °C and were harvested in parallel with the temperature shift larvae. RPA analysis of the indicated genes was then performed. “t” represents a tRNA control sample.
(B) As in (A), except that 5-d-old larvae were shifted from 29 °C to 21 °C, and controls remained at 29 °C.
All data are representative of at least three independent experiments.
To explore in more detail the mechanism underlying the changes in gene expression that follow temperature steps, we focused our attention on per4, a gene where we have previously studied the regulation by the circadian clock and light [27]. Do temperature steps affect the levels of per4 transcription? We investigated the effect of temperature steps on the expression of a 1.7-kilobase (kb) per4 promoter–luciferase reporter construct in stably transfected PAC-2 cells (1.7-kb wild-type [WT], Figure 4A) [27]. As a control for the direct effect of temperature on the rate of the luciferase-catalyzed bioluminescent reaction, in parallel we analyzed cells stably transfected with a luciferase reporter construct driven by the SV40 promoter and enhancer sequences (pGL3 Control). We previously confirmed that expression of this viral promoter reporter construct is neither regulated by the circadian clock nor by changes in temperature (unpublished data). Cells were exposed for 5 d in DD to 30 °C or 20 °C, then the temperature was decreased (30 °C to 20 °C, Figure 4B) or increased (20 °C to 30 °C, Figure 4C), and subsequently bioluminescence was assayed for 24 h. A temperature decrease resulted in sustained induction (3.34 ± 0.63-fold increase) (Figure 4B), while a temperature increase led to significant down-regulation of per4 promoter expression (5.40 ± 0.21-fold decrease) (Figure 4C) during the first 11 h following the temperature steps, relative to the pGL3 Control. Expression subsequently remained at these new levels for the remainder of the analysis (Figure 4B and 4C). Using RPA analysis, we confirmed that increases and decreases of luciferase mRNA expression occur following the temperature steps in the 1.7-kb WT transfected cells (Figure S2). These results reveal that a transcriptional mechanism is at least in part responsible for the changes in expression that follow temperature steps.
Figure 4 Temperature Steps Induce Changes in per4 Gene Transcription
(A) Schematic representation of the 1.7-kb WT (red), 0.4-kb WT (green), and the 0.4-kb Mut −7/−156/−173 (blue) per4 promoter luciferase reporter constructs. E-box elements are represented by rectangles (
CACGTG) and ellipses (
AACGTG), and their positions relative to the principal transcription start site are labeled. E-boxes that have been ablated by mutation to the sequence (
CTCGAG) are shown by a cross [27].
(B) Bioluminescence from PAC-2 cells stably transfected with the 1.7-kb WT construct adapted to 30 °C and then shifted rapidly to 20 °C (red trace). During the entire assay, the plate was held inside the Topcount counting chamber, and each well was counted for 3 s at intervals of approximately 5.5 min. Bioluminescence (counts per second) is plotted against time (h) following the temperature shift. A black trace represents pGL3Control transfected cells bioluminescence.
(C) Equivalent experiment to that in panel B, with cells adapted to 20 °C and shifted to 30 °C.
(D and E) Cells transfected with the 0.4-kb WT and 0.4-kb Mut −7/−156/−173 constructs were subjected to the same rapid temperature decrease and increase, respectively, as described for (B) and (C).
All traces represent the mean values of 16 independent wells. Each panel is representative of at least three independent experiments.
Do these temperature-dependent changes in transcription use the same pathway or regulatory elements that mediate the clock regulation of per4, i.e., the E-box enhancer elements? We have previously demonstrated that mutation of E-box elements within the per4 promoter eliminates circadian clock control [27]. Therefore, we compared the acute temperature response of a 0.4-kb per4 promoter luciferase reporter construct containing three E-box elements (0.4-kb WT, Figure 4A) with the same construct where the sequences of all E-box elements have been mutated [0.4-kb mutant (Mut) −7/−156/−173]. The mutation of the E-box elements did not eliminate the increases (1.5 ± 0.16-fold) or decreases (3.0 ± 0.11-fold) in expression following temperature steps that were observed in the 0.4-kb WT construct (1.9 ± 0.17-fold increase and 4.1 ± 0.12-fold decrease) (Figure 4D and 4E). Thus, the results reported here demonstrate that promoter elements distinct from E-boxes, within 0.4 kb of the per4 transcription start site, are responsible for driving the changes in expression in response to temperature steps.
The Zebrafish Circadian Clock Is Temperature-Compensated
An essential property of the zebrafish circadian clock is that it accurately measures time over a range of temperatures. Temperature-dependent changes in core clock function may be predicted to compensate for the tendency of the speed of the oscillator to vary with the ambient temperature [1,2]. We initially wished to confirm that the zebrafish clock is temperature-compensated. We measured the period length of circadian-clock-generated rhythms entrained by LD cycles as a function of temperature, in two luciferase reporter constructs. The first construct contained four E-box elements within their natural context of the per4 promoter (1.7-kb WT), while the second was a heterologous promoter containing four copies of a per4 E-box element cloned upstream of a synthetic
TATA box [4xE-box (−7)]. The reporter cells were assayed during incubation at a constant 20 °C, 25 °C, or 30 °C (Figure 5 and [27]) for 3 d in LD cycles and then transferred to DD for 3 d. We measured the period length of the rhythmic luciferase expression between the second and third day in DD. The free-running period length (τ) for the E-box heterologous promoter was 22.98 ± 0.17 h at 20 °C, 25.19 ± 0.21 h at 25 °C, and 25.46 ± 0.15 h at 30 °C, predicting that the temperature coefficient Q10 = 0.9 over the range 20 °C to 30 °C (see Materials and Methods, and Figure 5A and 5B). This confirms that the zebrafish clock indeed shows temperature compensation over a 10 °C range. However, calculations of period length were more problematic for the per4 promoter reporter, particularly at 20 °C, due to its very rapid dampening in DD conditions (Figure 5C and 5D). Interestingly, the bioluminescence traces of both constructs showed a pronounced change in rhythm amplitude between 20 °C (Figure 5A and 5C) and 30 °C (Figure 5B and 5D), both in LD and free running conditions [for 1.7-kb WT, 12.1 ± 1.6-fold at 30 °C, and 3.3 ± 0.56-fold at 20 °C, and for 4xE-box (−7) 9.7 ± 1.0-fold at 30 °C, and 1.6 ± 0.11-fold at 20 °C as measured on the second d in LD]. To confirm that this also occurs in the case of the endogenous per4 gene, we examined per4 RNA expression in cells entrained by an LD cycle at a constant 20 °C or 30 °C (Figure 5E and 5F). A higher-amplitude rhythm of expression was observed at 30 °C (7.9 ± 0.56-fold difference between peak and trough), compared with that at 20 °C (1.91 ± 0.5-fold). Thus, at least in part, the ambient temperature strongly influences the amplitude of circadian rhythms of transcription, during and following entrainment by LD cycles. Such a property has been already proposed by mathematical models to explain temperature compensation of the circadian clock [33–35].
Figure 5 Temperature Compensation and the Amplitude of E-box-Directed Rhythmic Expression
(A) Bioluminescence profile of 4xE-box (−7) reporter cells held at 20 °C under a LD cycle and then transferred to DD conditions. Plates were counted once per hour and maintained in robotic stacking units between assays, where they were illuminated.
(B) Equivalent experiment to panel A, with cells maintained at 30 °C.
(C) Bioluminescence traces from 1.7-kb WT per4 reporter cells maintained at 20 °C under LD cycle and DD conditions.
(D) Bioluminescence traces from 1.7-kb WT per4 reporter cells maintained at 30 °C under LD cycle and DD conditions.
(E) RPA analysis of per4 expression in WT PAC-2 cells held at 20 °C and 30 °C under an LD cycle for 3 d. RNA extracts were prepared on the fourth day at 3-h intervals during one 24-h cycle. Time 0 represents ZT 0: the onset of the light period. A white and black bar above the autoradiograph indicates the duration of the light and dark periods. RPA results with a β-actin loading control are also shown. “t” represents a tRNA control sample.
(F) A bar graph shows quantification of the peak (ZT3) and trough (ZT15) per4 expression values at 20 °C and 30 °C plotted as described in Figure 1, with error bars representing the standard deviation of three independent experiments.
All bioluminescence traces represent the mean values of 16 independent wells. Each panel is representative of at least three independent experiments.
E-box Function Is Influenced by Temperature
What makes the amplitude of E-box-driven rhythmic expression respond to temperature? The basic helix-loop-helix Per-Arnt-Sim domain clock proteins, CLK and BMAL, bind to E-boxes as heterodimers and thereby activate transcription at certain phases of the circadian clock cycle [13]. Thus, to test whether transcriptional activation mediated by CLK and BMAL was temperature dependent, we transiently transfected the 4xE-box (−7) luciferase reporter plasmid together with various combinations of zebrafish CLK1, 2, and 3, and BMAL1 and 2 expression constructs. We compared the levels of activation between cells incubated at 20 °C and 30 °C (Figure 6A). At 30 °C, activation driven by all three CLK family members alone was 5- to 6-fold higher than at 20 °C. Co-transfection of CLK1 and CLK3 in combination with BMAL1 led to an even higher activation at 30 °C relative to 20 °C (6- to 10-fold). In contrast, the activation obtained by transfecting BMAL 1 or 2, the reporter construct alone, or the pGL3Control plasmid was only 1.5- to 2-fold higher at 30 °C than at 20 °C. Thus, the amplitude of transcriptional activation driven by the members of the CLK family, alone or in combination with BMAL1, appears to be strongly temperature-dependent in zebrafish cells.
Figure 6 Temperature Influences CLK Protein Expression and Function
(A) In vitro luciferase assays of transiently transfected PAC2 cells. The combinations of CLK (Clk) and BMAL (Bml) expression vectors cotransfected with the 4x Ebox (−7) reporter plasmid are indicated for each assay result. Control cells were transfected with the reporter plasmid or with the pGL3 Control plasmid alone. Values represent the mean fold difference between luciferase activities measured in 30 °C and 20 °C, 60 h after transfection. All assays were standardized for transfection efficiency using a β-galactosidase assay. The results are based on four independent experiments, and error bars indicate the standard deviation.
(B) Electrophoretic mobility shift assay of nuclear extracts from PAC-2 cells cultured at 20 °C or 30 °C on a LD cycle, and harvested at ZT3, 9, 15, and 21 (lanes 1 to 8). Three specific complexes are indicated by A, B, and an asterisk. Supershift assays of a ZT15, 30 °C extract (+Ab), used either a dopamine transporter antibody (Control) or a mouse clk antibody (Clock) (lanes 9 and 10). Complexes indicated by A, B, and an asterisk are all efficiently competed by a 25-, 50-, and 100-fold excess of cold E-box probe (lanes 12, 13, and 14, respectively, and compare with lane 11), but not with a 100-fold excess of a CRE probe (compare lane 15 with lane 11).
(C) Western blotting assay using the anti-mouse CLK antibody of the same nuclear extracts tested in the electrophoretic mobility shift assay analysis of panel B. The migration of a 100-kDa marker band is shown. Below are shown western blotting results for the same extracts using an anti-mouse CREB antibody as a loading control.
(D) Western blot assay of CLK protein in 30 °C extracts prepared at ZT9 or ZT21 (time points representing the trough and peak, respectively, of the CLK protein rhythm). Samples were prepared with (+) or without (−) treatment with alkaline phosphatase prior to electrophoresis and transfer. In panels B, C, and D, data are representative of at least three independent experiments.
Does temperature influence the binding of endogenous CLK protein-containing complexes to circadian E-boxes? Nuclear extracts were prepared at specific time points from cells maintained under a LD cycle, at 20 °C or 30 °C and then tested for binding to an E-box probe containing two consensus E-box sequences, by electrophoretic mobility shift assay (Figure 6B). Interestingly, levels of two slow mobility complexes varied according to the temperature and the time of day. Levels of one complex (complex A) in 30 °C cells followed a 24-h rhythm with peak levels at Zeitgeber Time (ZT) ZT15 and a trough at ZT3 (Figure 6B, lanes 6 and 2, respectively). In contrast, at 20 °C this complex was barely detectable at all time points. At 30 °C, levels of a second complex (complex B) showed a rhythm similar to that observed for complex A, while at 20 °C peak levels were reduced and shifted to ZT21 (lanes 1, 3, 5, and 7). Levels of a third abundant complex (indicated by an asterisk) did not change significantly according to the time of day or temperature. In control experiments, the binding of all three complexes at ZT15 at 30 °C were efficiently competed by an excess of cold E-box sequence (lanes 11–14) but not by an unrelated cAMP response element sequence (lane 15). Thus, it appears that temperature affects the levels of certain specific E-box-binding complexes. In order to test whether CLK proteins are components of these complexes, we performed a supershift assay using a mouse CLK-specific antibody (Figure 6B). This antibody efficiently supershifted both complexes A and B, but not the abundant complex (indicated by an asterisk) (lane 10), while a control antibody (anti-dopamine transporter) failed to supershift any of the complexes (lane 9). To test which of the three zebrafish CLK proteins were recognized by this mouse antibody, myc-tagged versions of the CLK proteins were expressed in an in vitro translation system and were then analyzed by western blotting. While all zebrafish CLK proteins were efficiently recognized by the control myc tag antibody, only CLK1 (and, to a lesser extent, CLK3) were detected by the mouse CLK-specific antibody (Figure S3).
Do the levels of CLK protein change at the high and low temperatures? We performed a western blot analysis of the same nuclear extracts prepared for the electrophoretic mobility shift assay using the mouse CLK-specific antibody. This assay revealed, at all four (20 °C) time points, the presence of two closely spaced bands, similar in size to that predicted for the CLK1 protein (approximately 100 kDa, Figure 6C). The overall levels of CLK-immunoreactive bands were comparable at the four time points; however, at ZT21 the two bands showed an equal intensity while at ZT9 the high molecular weight band was stronger. In contrast, at 30 °C, overall CLK protein levels changed during the 24-h cycle, with peak levels at ZT21 and a trough at ZT9 (Figure 6C). Also, at 30 °C, the relative levels of the two bands changed considerably between the four time points. Specifically, the intensity of the higher molecular weight band decreased significantly between ZT9 and ZT21, while the intensity of the lower molecular weight band increased (Figure 6C).
Does this result reflect temperature-dependent changes in CLK post-translational modifications? The phosphorylation status of several clock proteins in various model systems has been shown to vary through the circadian cycle, a property linked with changes in the protein stability or function [36–38]. In many cases, different levels of phosphorylation can be visualized by changes in electrophoretic mobility of the protein. We tested whether the multiple CLK-immunoreactive bands represented various phosphorylated forms of this protein. Extracts from the peak (ZT21) and trough (ZT9) points in the 30 °C cell extracts were treated with phosphatase before western analysis in parallel with untreated controls. Phosphatase treatment of the ZT9 extract increased the mobility of the CLK band so that it co-migrated with the single CLK-immunoreactive band in the ZT21 extract (Figure 6D). The mobility of the ZT21 extract band did not alter with phosphatase treatment. This result points to temperature affecting the amplitude of cycling CLK protein levels and their phosphorylation.
Discussion
The zebrafish, as an ectotherm, represents an ideal vertebrate model system to study the effects of temperature on circadian clock function. Given the geographical distribution, and the shallow, fresh-water habitats of natural populations of zebrafish (data from http://www.fishbase.org, it seems likely that their core body temperature would naturally be subjected to a day–night rhythm, as well as seasonal changes. It is therefore reasonable to predict that temperature would normally play a role in regulating the circadian timing system. Our results have highlighted four responses of the zebrafish clock to temperature: (1) entrainment by even shallow temperature cycles, (2) regulation of the expression levels of many clock genes by temperature steps in a gene-specific manner, (3) temperature compensation, and (4) a strong effect of the ambient temperature on the amplitude of cycling expression of certain clock genes. Furthermore, our observation that the clock in cell lines responds directly to temperature reinforces the notion of autonomy in zebrafish peripheral clock entrainment and provides a valuable cell culture tool to explore the temperature response [22,27].
Molecular Mechanisms for Entrainment by Temperature
Our results clearly point to temperature cycles entraining rhythms of circadian gene expression in zebrafish. We base this conclusion on two observations: (i) per4 expression rhythms persist after transfer from temperature cycles to constant temperature, and (ii) the phase of the per4 rhythm changes relative to the phase of the temperature cycle as a function of the length of the thermoperiod. However, the response of the per4 rhythm to temperature cycles with different thermoperiods suggests that temperature changes could also drive expression of certain clock genes (the increase in per4 expression consistently coincides with the warm–cold transition). Indeed, we show that acute temperature steps significantly alter the expression levels of several clock genes in a gene-specific manner, not simply reflecting global changes in transcription rate. In the case of the per4 promoter, these changes do not involve circadian clock regulation via E-boxes, suggesting a temperature-driven response (see scheme in Figure 7A). The acute responses observed for per4, cry3, and cry2a after the temperature steps match their expression profile, following the individual temperature transitions under temperature cycle conditions. Thus, we speculate that the acute temperature regulation of these genes may contribute to entrainment of the clock mechanism and so represents a component of the temperature input pathway (Figure 7A). Within this input pathway, the gene expression response may lie downstream of more rapid temperature-dependent regulatory events that also contribute to entrainment of the clock. The per4 gene expression rhythm observed under temperature cycles may represent the integration of the temperature-driven response (via element X) and regulation by the entrained circadian clock (via E-boxes). Further studies will be required to test these hypotheses.
Figure 7 Model for Temperature Regulation of the per4 Promoter
(A) Temperature steps entrain the phase of the clock by driving expression levels of per4 and other clock genes via a hypothetical enhancer element X. Temperature decreases result in expression increases, and vice versa. Although E-boxes ultimately mediate regulation of the per4 promoter by the entrained clock, they do not participate in the temperature-driven response.
(B) Temperature influences the amplitude of rhythmic per4 expression that has been entrained by LD cycles in two ways: (1) by determining the amplitude of E-box-directed rhythmic expression, via changes in CLK protein levels, phosphorylation, and E-box binding, and (2) by driving expression changes through element X (see panel A). The promoter integrates these two regulatory mechanisms. The temperature-dependent amplitude of E-box-directed rhythmic expression would be predicted to involve the core feedback loops of the clock itself and, according to mathematical models, might thereby underlie temperature compensation.
An ectothermic organism needs to modify many aspects of its physiology in order to adapt to substantial changes in core body temperature. The presence of circadian and circannual clocks and the ability to respond to photoperiodic changes provide mechanisms to anticipate regular daily and seasonal temperature changes and so give sufficient time to mount an appropriate gene expression response [39]. However, many studies have shown that changes in gene expression also occur at the cellular level in direct response to temperature alterations [39–41]. The transcriptional mechanism that regulates per4 expression following temperature shifts may therefore constitute a more general mechanism whereby cells perceive and adapt to temperature changes.
By comparing gene expression under LD and temperature cycle conditions, we have confirmed previous reports that per2 expression is light driven [30,31] and now show that this clock gene is not induced by all clock-entraining signals (zeitgebers). Thus, light and temperature cycles appear, at least in part, to drive gene expression within the circadian clock by distinct pathways. The presence of larger clock gene families in teleosts may have led to specialization of individual genes to respond to single zeitgebers [25].
Temperature Regulates the Amplitude of Rhythmic Clock Gene Expression: A Mechanism Underlying Temperature Compensation?
Temperature compensation of the circadian clock is essential to preserve its timing function over a range of temperatures [1]. We have confirmed that the zebrafish PAC-2 cell clock is temperature-compensated, actually decreasing the rate of its oscillation slightly when the temperature is increased, as has been reported for other cell culture model systems [7,8]. Mathematical models in which temperature influences the amplitude of the circadian pacemaker have been proposed to explain various aspects of the behavior of circadian clock outputs including their temperature compensation [33–35]. Consistently, we observe that the amplitude of circadian E-box-directed rhythmic transcription entrained by LD cycles is 6-fold higher at 30 °C than at 20 °C, while the phase remains constant. The temperature also influences the amplitude of per4 rhythmic expression. However, the changes in the amplitude of the E-box heterologous promoter rhythm originate from differences in the peak values, while for the per4 promoter, the amplitude is determined by differences in the trough values. To explain this apparent discrepancy, we propose a model (Figure 7B) where the per4 promoter can integrate temperature and light regulatory input from the E-boxes together with regulation by other temperature-driven elements (X in our model). This ultimately results in the trough levels of expression being set by the E-box-independent, temperature-driven regulation (higher trough levels at lower temperatures). Instead, the relative levels of the peaks respond to E-box input (peak values remain constant since the amplitude of light-cycle-entrained rhythmic expression decreases at lower temperatures) (Figure7B).
We have explored the mechanism whereby the amplitude of E-box-directed expression rhythms respond to the temperature. The CLK-directed activation of an E-box reporter construct is 5- to 10-fold higher at 30 °C than at 20 °C as measured in transfection assays. This strong effect of temperature on transcriptional activation appears to be CLK-specific, because not only is E-box reporter expression driven by BMAL alone, expression of control reporter constructs differs by only 1.5- to 2-fold over the same 10 °C range. Our studies using PAC-2 cells have also revealed that the levels of endogenous CLK-containing nuclear complexes that bind specifically to E-boxes as well as CLK protein levels and phosphorylation change as a function of temperature and time of day during entrainment by a LD cycle. We speculate that these properties may ultimately define the E-box rhythm amplitude (Figure 7B). Indeed, regulation of the phosphorylation of clock proteins has been tightly linked with other basic properties of the circadian clock [36–38].
Looking from a broader perspective, the temperature-responsive transcriptional regulatory mechanisms that we have revealed in this study may form part of more general mechanisms that directly adapt gene expression and cell physiology to changes in ambient temperature. Thus the implications of our work may reach beyond the circadian clock.
Materials and Methods
RNA and protein analysis
RNA extractions from larvae and cells, RPAs, and the per4, β-actin, and clock1 riboprobes have been described [27,28]. The remaining riboprobes were for per2 [26,42] (from positions 3113–3874 relative to the translation start site), for cry2a [29] (positions 1589–1968), and for cry3 [29] (positions 1339–1797). The clock cDNAs were transcribed and translated using the TnT-Quick Coupled Transcription/Translation System (Promega, Madison, Wisconsin, United States), before western blotting (BioRad, Hercules, California, United States) using an anti-mouse CLK (Santa Cruz Biotechnology, Santa Cruz, California, United States) or myc antibody (Upstate Biotechnology, Lake Placid, New York, United States), and visualization with the ECL detection system (Amersham Biosciences, Little Chalfont, United Kingdom). For phosphatase treatments, nuclear extracts were prepared as described for the electrophoretic mobility shift assays, without the addition of phosphatase inhibitors. Extracts were then treated with 1 unit of calf intestinal phosphatase (Roche, Basel, Switzerland) in nuclear extract buffer, at 37 °C for 15 min. Laemmli buffer was added to a final concentration of 1× and the samples heated at 95 °C for 5 min before SDS electrophoresis and western blotting analysis. The polyclonal anti-mouse CREB antibody was purchased from Cell Signaling Technology (Beverly, Massachusetts, United States).
Transient transfection assays
A standard electroporation method was used [27]. Transfected cells were assayed using an in vitro luciferase assay system (Promega). The CLK and BMAL expression constructs were based on the pcDNA3.1/Myc-His expression vector (Invitrogen, Carlsbad, California, United States). Co-transfection with the plasmid pcDNA3.1/Myc-His (+)/lacZ and a β-galactosidase assay served to control for transfection efficiency.
Electrophoretic mobility shift and supershift assays
Preparation of nuclear extracts, radioactive labeling and purification of oligonucleotide probes, and EMS assays were performed as described elsewhere [43]. The E-box probes contained a tandem repeat of the per4 promoter E-box (−7) sequence (sense oligo: 5′-
GAAGCACGTGTACTCGGAAGCACGTGTACTCG-3′) [27]. Supershift assays using an anti-mouse CLK and dopamine transporter antibodies (Santa Cruz Biotechnology) were performed as described elsewhere [43].
Cell cultures and in vivo luciferase assays
Culture conditions and in vivo luciferase assays have been described [27]. In 0.4-kb Mut −7/−156/−172 contains a per4 promoter fragment, extending between −207 and +190 relative to the transcription start site, cloned in pGL3Basic where the E-box sequences are mutated to
CTCGAG by site-directed mutagenesis.
Raising zebrafish larvae, temperature, and lighting control
The zebrafish Tübingen strain was maintained and crossed using standard methods [20]. Flasks containing cells or larvae were submerged in 60-l water baths with circulating heating and cooling units (Lauda, Lauda-Königshofen, Germany) and illuminated with a tungsten light source (11 μW/cm2) [24]. Temperature cycles were generated by controlling the heating and cooling units using Wintherm plus software (Lauda).
Data analysis
Bioluminescence data were analyzed using Microsoft Excel or CHRONO software [27,44]. Period estimates measured after 2 d in DD were made by linear regression following peakfinder analysis with CHRONO [44]. For Q10 temperature coefficient calculations, period length estimates for cells held at 20 °C, 25 °C, and 30 °C were calculated as cycles per hour and then plotted against temperature. Linear regression analysis revealed a good fit to a straight line (R2 = 0.9734). Mean period lengths at 20 °C and 30 °C were then substituted into the equation Q10 = (R2/R1)10/(T2-T1), where R is rate and T is temperature. Autoradiographic images were quantified with the aid of Scion Image software (NIH, http://rsb.info.nih.gov/nih-image/). Statistical analysis was performed with the aid of GraphPad PRISM 4 software (GraphPad Software, San Diego California, United States).
Supporting Information
Figure S1 Temperature Cycles Induce Rhythmic per4 mRNA Expression in Larvae and PAC-2 Cells
(A) RPA analysis of per4 and β-actin expression in larvae raised for 7 d in DD on a 2 °C temperature cycle (24 °C/11.5 h, 26 °C /11.5 h, plus an additional 0.5 h for both heating and cooling phases). During the seventh day, RNA was harvested at the indicated times (Time 0 is defined as the beginning of the heating period).
(B) RPA analysis of per4 and β-actin expression in PAC-2 cells cultured for 7 d in a 4 °C temperature cycle (23.5 °C/11 h, 27.5 °C/11 h, plus 1 h for each heating and cooling phase) under DD conditions. Cells were harvested during the seventh day. Control cells maintained at a constant temperature (25 °C) during the entire experiment were harvested and assayed in parallel.
(2.16 MB TIF).
Click here for additional data file.
Figure S2 Temperature Steps Induce Changes in Luciferase Reporter mRNA Expression
(A) Quantification of RPA analysis of luciferase mRNA expression in 1.7-kb WT stably transfected cells, 0, 1, 3, and 6 h following transfer from 30 °C to 20 °C. The experiment was performed in triplicate, and error bars denote the standard deviation.
(B) Equivalent analysis of luciferase expression in cells transferred from 20 °C to 30 °C.
(512 KB TIF).
Click here for additional data file.
Figure S3 Analysis of Recombinant Zebrafish CLOCK Proteins
Western blotting analysis of in vitro transcription/translation extracts containing myc-tagged CLOCK proteins (Clock-myc 1, 2, and 3). Blots were treated with an anti-myc tag monoclonal antibody (myc-Ab) or an anti-mouse CLK polyclonal antibody (Clock-Ab).
(2.24 MB TIF).
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 clock1 (AF133306), clock2 (AB087255), clock3 (AB087256), cry2a (AB042250), cry3 (AB042252), and per2 (AY171100).
We would like to thank Ferenc Müller, Darren Gilmour, David Whitmore, Andreas Heyd, and Cristiano Bertolucci for critical reading and very helpful comments. We are also extremely grateful for the excellent technical assistance of Andreas Heyd. This work was supported by funds and fellowships from the Max -Planck Society and the Centre National de la Recherche Scientifique (CNRS: France). NSF participated in a joint CNRS/Max-Planck Society exchange program. TD was supported by a long-term fellowship from the European Molecular Biology Organization.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. KL and NSF conceived and designed the experiments. KL and CS performed the experiments. KL, DV, and NSF analyzed the data. DV, SBG, and CS contributed reagents/materials/analysis tools. DV, TD, and NSF wrote the paper.
Citation: Lahiri K, Vallone D, Gondi SB, Santoriello C, Dickmeis T, et al (2005) Temperature regulates transcription in the zebrafish circadian clock. PLoS Biol 3(11): e351.
Abbreviations
BMALBrain and muscle Arnt-like protein
CLKclock
Crycryptochrome
DDconstant darkness
kbkilobase
LDlight–dark
Mutmutant
Perperiod
RPARNase protection assay
WTwild-type
ZTZeitgeber Time
==== Refs
References
Sweeney BM Hastings JW Effects of temperature upon diurnal rhythms Cold Spring Harb Symp Quant Biol 1960 25 87 104 13774256
Pittendrigh CS On temperature independence in the clock system controlling emergence time in Drosophila
Proc Natl Acad Sci U S A 1954 40 1018 1029 16589583
Liu Y Garceau NY Loros JJ Dunlap JC Thermally regulated translational control of FRQ mediates aspects of temperature responses in the neurospora circadian clock Cell 1997 89 477 486 9150147
Martino-Catt S Ort DR Low temperature interrupts circadian regulation of transcriptional activity in chilling-sensitive plants Proc Natl Acad Sci U S A 1992 89 3731 3735 1570291
Pittendrigh CS Temporal organization: Reflections of a Darwinian clock-watcher Annu Rev Physiol 1993 55 16 54 8466172
Underwood H Calaban M Pineal melatonin rhythms in the lizard Anolis carolinensis : I. Response to light and temperature cycles J Biol Rhythms 1987 2 179 193 2979659
Izumo M Johnson CH Yamazaki S Circadian gene expression in mammalian fibroblasts revealed by real-time luminescence reporting: Temperature compensation and damping Proc Natl Acad Sci U S A 2003 100 16089 16094 14657355
Tsuchiya Y Akashi M Nishida E Temperature compensation and temperature resetting of circadian rhythms in mammalian cultured fibroblasts Genes Cells 2003 8 713 720 12875656
Ruby NF Burns DE Heller HC Circadian rhythms in the suprachiasmatic nucleus are temperature-compensated and phase-shifted by heat pulses in vitro J Neurosci 1999 19 8630 8636 10493763
Damiola F Le Minh N Preitner N Kornmann B Fleury-Olela F Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus Genes Dev 2000 14 2950 2961 11114885
Brown SA Zumbrunn G Fleury-Olela F Preitner N Schibler U Rhythms of mammalian body temperature can sustain peripheral circadian clocks Curr Biol 2002 12 1574 1583 12372249
Wager-Smith K Kay SA Circadian rhythm genetics: From flies to mice to humans Nat Genet 2000 26 23 27 10973243
Reppert SM Weaver DR Molecular analysis of mammalian circadian rhythms Annu Rev Physiol 2001 63 647 676 11181971
Liu Y Merrow M Loros JJ Dunlap JC How temperature changes reset a circadian oscillator Science 1998 281 825 829 9694654
Majercak J Sidote D Hardin PE Edery I How a circadian clock adapts to seasonal decreases in temperature and d length Neuron 1999 24 219 230 10677039
Collins BH Rosato E Kyriacou CP Seasonal behavior in Drosophila melanogaster requires the photoreceptors, the circadian clock, and phospholipase C Proc Natl Acad Sci U S A 2004 101 1945 1950 14766972
Majercak J Chen WF Edery I Splicing of the period gene 3'-terminal intron is regulated by light, circadian clock factors, and phospholipase C Mol Cell Biol 2004 24 3359 3372 15060157
Sidote D Majercak J Parikh V Edery I Differential effects of light and heat on the Drosophila circadian clock proteins PER and TIM Mol Cell Biol 1998 18 2004 2013 9528772
Gekakis N Saez L Delahaye-Brown AM Myers MP Sehgal A Isolation of timeless by PER protein interaction: Defective interaction between timeless protein and long-period mutant PERL Science 1995 270 811 815 7481773
Westerfield M The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio) 2000 Eugene (Oregon) University of Oregon Press
Njus D McMurry L Hastings JW Conditionality of circadian rhythmicity: Synergistic action of light and temperature J Comp Physiol B 1977 117 335 344
Whitmore D Foulkes NS Sassone-Corsi P Light acts directly on organs and cells in culture to set the vertebrate circadian clock Nature 2000 404 87 91 10716448
Kaneko M Cahill GM Light-dependent development of circadian gene expression in transgenic zebrafish PLoS Biol 2005 3 e34 15685291
Dekens MP Santoriello C Vallone D Grassi G Whitmore D Light regulates the cell cycle in zebrafish Curr Biol 2003 13 2051 2057 14653994
Cahill GM Clock mechanisms in zebrafish Cell Tissue Res 2002 309 27 34 12111534
Delaunay F Thisse C Thisse B Laudet V Differential regulation of Period 2 and Period 3 expression during development of the zebrafish circadian clock Gene Expr Patterns 2003 3 319 324 12799078
Vallone D Gondi SB Whitmore D Foulkes NS E-box function in a period gene repressed by light Proc Natl Acad Sci U S A 2004 101 4106 4111 15024110
Whitmore D Foulkes NS Strahle U Sassone-Corsi P Zebrafish clock rhythmic expression reveals independent peripheral circadian oscillators Nat Neurosci 1998 1 701 707 10196586
Kobayashi Y Ishikawa T Hirayama J Daiyasu H Kanai S Molecular analysis of zebrafish photolyase/cryptochrome family: Two types of cryptochromes present in zebrafish Genes Cells 2000 5 725 738 10971654
Pando MP Pinchak AB Cermakian N Sassone-Corsi P A cell-based system that recapitulates the dynamic light-dependent regulation of the vertebrate clock Proc Natl Acad Sci U S A 2001 98 10178 10183 11517315
Cermakian N Pando MP Thompson CL Pinchak AB Selby CP Light induction of a vertebrate clock gene involves signaling through blue-light receptors and MAP kinases Curr Biol 2002 12 844 848 12015122
Pittendrigh CS Circadian rhythms and the circadian organization of living systems Cold Spring Harb Symp Quant Biol 1960 25 159 184 13736116
Pittendrigh CS Kyner WT Takamura T The amplitude of circadian oscillations: Temperature dependence, latitudinal clines, and the photoperiodic time measurement J Biol Rhythms 1991 6 299 313 1773097
Lakin-Thomas PL Brody S Cote GG Amplitude model for the effects of mutations and temperature on period and phase resetting of the Neurospora circadian oscillator J Biol Rhythms 1991 6 281 297 1837742
Barrett RK Takahashi JS Lability of circadian pacemaker amplitude in chick pineal cells: A temperature-dependent process J Biol Rhythms 1997 12 309 318 9438879
Liu Y Loros J Dunlap JC Phosphorylation of the Neurospora clock protein FREQUENCY determines its degradation rate and strongly influences the period length of the circadian clock Proc Natl Acad Sci U S A 2000 97 234 239 10618401
Nakajima M Imai K Ito H Nishiwaki T Murayama Y Reconstitution of circadian oscillation of cyanobacterial KaiC phosphorylation in vitro Science 2005 308 414 415 15831759
Sathyanarayanan S Zheng X Xiao R Sehgal A Post-translational regulation of Drosophila PERIOD protein by protein phosphatase 2A Cell 2004 116 603 615 14980226
Johnston IA Cold adaptation in marine organisms Philos Trans R Soc Lond B Biol Sci 1990 326 655 666 discussion 666-667 1969650
Airaksinen S Jokilehto T Rabergh CM Nikinmaa M Heat- and cold-inducible regulation of HSP70 expression in zebrafish ZF4 cells Comp Biochem Physiol B Biochem Mol Biol 2003 136 275 282 14529753
Imamura S Ojima N Yamashita M Cold-inducible expression of the cell division cycle gene CDC48 and its promotion of cell proliferation during cold acclimation in zebrafish cells FEBS Lett 2003 549 14 20 12914916
Hirayama J Fukuda I Ishikawa T Kobayashi Y Todo T New role of zCRY and zPER2 as regulators of sub-cellular distributions of zCLOCK and zBMAL proteins Nucleic Acids Res 2003 31 935 943 12560489
Vallone D Pellecchia MT Morelli M Verde P DiChiara G Behavioural sensitization in 6-hydroxydopamine-lesioned rats is related to compositional changes of the AP-1 transcription factor: Evidence for induction of FosB- and JunD-related proteins Brain Res Mol Brain Res 1997 52 307 317 9495553
Roenneberg T Taylor W Automated recordings of bioluminescence with special reference to the analysis of circadian rhythms Methods Enzymol 2000 305 104 119 10812594
|
16176122
|
PMC1233578
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Nov 27; 3(11):e351
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030351
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030379SynopsisCell BiologyMolecular Biology/Structural BiologyNeuroscienceDanio (Zebrafish)Temperature Regulates the Zebrafish Clock Synopsis11 2005 27 9 2005 27 9 2005 3 11 e379Copyright: © 2005 Public Library of Science.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.
Temperature Regulates Transcription in the Zebrafish Circadian Clock
==== Body
The conveniences of modernity are not without their costs, as weary travelers know all too well. When you jet into a country halfway around the world and it's dark hours before your body expects nighttime, your internal clock doesn't have time to adjust, so you feel jet-lagged. The biological clock, calibrated to daily light and temperature cycles, controls the circadian rhythms of a wide range of physiological and behavioral processes, from fluctuating hormone levels to sleep–wake cycles and feeding patterns. While it's well known that circadian clock elements sense and respond to light cycles, evidence of temperature-dependent changes in multiple cellular processes—such as gene transcription, translation, and protein stability—in fruit flies and fungi suggest that many circadian clock components also respond to temperature cycles.
Daily temperature cycles and spikes can reset the clock's phase (timing of the peaks and troughs of activity), though its cycle length remains fixed over a wide range of temperatures. This “temperature compensation” feature confers a measure of resistance to the potentially disruptive effects of temperature fluctuations on the accuracy of the clock's timing mechanism, and appears to be a general property of the circadian system.
Since little is known about how vertebrates manage these temperature-related responses at the genetic and molecular level, Kajori Lahiri, Nicholas Foulkes, and their colleagues decided to study this question in zebrafish. This genetically tractable model organism is especially suited to this task, the authors explain, because adults, larvae, and even embryos can tolerate a wide range of core body temperatures (being cold-blooded animals) that can be manipulated simply by changing the water temperature. Temperature variations of as little as 2 °C (35.6 °F) can reset the zebrafish clock, Lahiri et al. show, and precise shifts in temperature trigger significant changes in the expression of specific clock genes.
To test whether temperature cycles can establish, or entrain, circadian rhythms in zebrafish like light–dark cycles do, Lahiri et al. raised zebrafish larvae in total darkness for six days, starting four hours after fertilization, and exposed them to a 4 °C temperature cycle. A subset of fish (at the larval stage) were sacrificed every three hours to measure RNA levels of core clock genes (per2, per4, cry2a, cry3, and clock1) and determine their expression profiles. As a control, sibling larvae were exposed to light–dark cycles and constant temperature. Clock genes per4, cry2a, cry3, and clock1 showed rhythmic expression under both light–dark and temperature cycles, with the high temperature phase matching the light phase. Remarkably, the authors wrote, the results were similar for larvae raised with a daily temperature fluctuation cycle of as little as 2 °C.
Zebrafish cell lines also proved valuable tools for studying temperature response, showing a similar pattern of clock gene expression under a 4 °C temperature–darkness cycle as the larvae did under a 2 °C temperature–darkness cycle. Expression of per4 continued even after the cells were exposed to constant temperature, an indicator of entrainment. Temperature shifts can also trigger significant changes in clock gene expression (transcript levels of per4 and cry3 dropped after a temperature increase and spiked after a temperature decrease; cry2 showed the opposite response)—changes wrought by temperature-dependent shifts in the behavior of transcriptional regulators, as in the case of per4. Acute shifts in temperature alter the expression of several clock genes selectively. How these gene-expression responses fit into this temperature-triggered pathway is unclear, but Lahiri et al. offer a few hypotheses for future testing—investigations that should benefit from using the zebrafish cell lines the authors developed.
Several key transcriptional regulatory elements at the core of the zebrafish circadian clock respond to temperature
The authors go on to show that the zebrafish clock functions over a range of temperatures with characteristic temperature compensation. They speculate that this may result when temperature changes produce shifts in the amplitude of circadian transcription rhythms. Altogether these results show that temperature can regulate the circadian clock in this vertebrate. If the temperature-induced transcriptional responses described here operate in other temperature-related responses, they may shed light on how temperature affects other biological systems as well. —Liza Gross
|
0
|
PMC1233579
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Nov 27; 3(11):e379
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030379
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1616784510.1371/journal.pmed.0020337Policy ForumInfectious DiseasesOtherBiocrimes, Microbial Forensics, and the Physician Policy ForumSchutzer Steven E *Budowle Bruce Atlas Ronald M Steven E. Schutzer is a physician-scientist in the Department of Medicine, New Jersey Medical School, University of Medicine and Dentistry of New Jersey, Newark, New Jersey, United States of America. Bruce Budowle is Senior Scientist in the Laboratory Division, Federal Bureau of Investigation, Quantico, Virginia, United States of America. Ronald M. Atlas is in the graduate school of University of Louisville, Louisville, Kentucky, United States of America.
Competing Interests: The authors declare that no competing interests exist.
*To whom correspondence should be addressed. E-mail: [email protected] 2005 27 9 2005 2 12 e337Copyright: © 2005 Schutzer 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.Schutzer and colleagues give guidance for physicians who believe that one of their patients has been a victim of an act of bioterror or another biocrime.
==== Body
Attending physicians, regardless of their specialty or setting of practice, may suspect or learn that their patient has been attacked with a biological agent. In such cases, it is important to be aware of the interactions that may occur with law enforcement, and of the role that the evolving science of microbial forensics [1] may play in the investigation. Physicians are in key positions to preserve critical evidence and, thereby, contribute to the chain of custody, and, at the same time, offer suggestions to help develop the field of microbial forensics.
This article provides guidance to physicians who believe that one of their patients is a victim of an act of bioterrorism or of another biocrime, and who are compelled by law, or with the patient's consent, wish to assist law enforcement in an investigation. In this regard, there are instructive lessons that can be learned from cases of past biocrimes and from analogies to more familiar cases of sexual assault and child abuse (Table 1).
Table 1 Comparison of Biocrimes, Sexual Assaults, and Child Abuse in the US
It is recommended that current individual state requirements be consulted at the time of a particular occurrence.
Recent Biocrimes
The most widely publicized bioterrorism event in the United States was the US anthrax mail attacks of 2001. In this case, an astute physician diagnosed the index case of systemic anthrax [2] that set off national panic and a federal investigation that is still ongoing. Highly publicized in Europe was the assassination of a Bulgarian exile in London using ricin, a toxin extracted from castor beans, [3,4] which was delivered to him using an umbrella. Less publicized are other biocrimes such as the case of a laboratory worker in Texas intentionally infecting hospital co-workers with Shigella dysenteriae. Biocrimes are much less likely to occur than many infectious diseases, such as HIV/AIDS. In fact, it is often the abundance of naturally occurring infections that may make the detection of a biocrime difficult. However, there are also documented cases of non-bioterrorism biocrimes (Box 1) [5,6], which far exceed the number of documented bioterrorism acts, as well as many hoaxes where physicians and clinical laboratories were involved in determining if there was a real threat to exposed individuals.
Box 1. Examples of Non-Bioterror Biocrimes
Intentional Salmonella typhi food contamination in France from 1910 to1918
A Yersinia pestis attack by injection in 1933 in India
Deliberate use of HIV-infected blood and secretions to inflict harm
Source: [5,6].
(Reprinted cover image from [30] with permission from Elsevier)
A biocrime is similar to an assault crime, except, instead of a gun or knife, the weapon is a pathogen or a toxin. In the US, acts of bioterrorism are federal crimes that are governed by different responses by law enforcement and public health agencies than those that govern other biocrimes [7]. Most biocrimes and their subset of bioterrorism cases will involve public health agencies because of the nature of a disease threat to the public.
The numerous hoaxes that are biocrimes include white powders found in letters that proclaim the presence of anthrax, and threatening notes claiming ricin contamination of baby food. Ricin currently appears to be a prevalent bioweapon, particularly as a tool for extortion. These potential ricin threats demonstrate the impact of bioterrorism on patient care: physicians had to monitor patients who might have ingested the poisoned food and, hence, were distracted from caring for other patients [8]. Hoaxes can be challenging for the physician, who must distinguish between symptoms and signs that could be toxin-related and those that are just variants of normal health. This challenge was compounded in a recent case where there were trace findings of an inactive form of ricin in baby food. Nonetheless, a hoax is also a crime, and the physician should not discard any evidence simply because material appears innocuous.
As environmental biothreat sensors expand into the workplace and public places, they will be relied on as sentinels for possible biothreat releases. In the US, sensors for anthrax are being placed in some postal offices and at some major public events, and air monitoring is being carried out in major cities and transit systems (the BioWatch system). When such sensors indicate a possible bioterrorist attack, even if the signal is later found to be a false signal, the public will react, and are likely to seek out their own physicians for medical diagnostics, therapy, and advice. Potential cases that follow a public alert warrant evaluation, collection of patient samples, and possibly institution of prophylactic treatments until an alert is deemed a false alarm. Such a situation occurred in March 2005 following a presumptive positive detection of anthrax in a US Department of Defense mailroom. This situation necessitated treating over 900 potential victims with antibiotics as a precaution. However, despite the newer sensor technologies, physicians will likely remain the best and definitive authorities on the presence of an infection.
Reporting a Biocrime
Patients who believe that they have been a victim of a biocrime generally want both a medical and law enforcement response—that is, they want medical treatment, and they want the perpetrator to be found, prosecuted, and punished. Requirements for how physicians and other members of the health-care system initially report a suspected biocrime are governed at the local level (but regulations are ever-changing, and it is important to check current information at the time it is needed). Many US state regulations mandate that diagnostic laboratories report preliminary isolates of certain microbes to public health authorities, who, in turn, are sometimes compelled to notify law enforcement, depending on the isolate. Certain states have laws that mandate that a physician report specific diseases or unusual clinical manifestations to public health authorities and, in some cases, directly to law enforcement. The latter requirement is analogous to what is expected when there is suspicion of child abuse or gunshot wounds, and failure to notify authorities is itself a crime. The regulations of the Health Insurance Portability and Accountability Act (HIPAA) of 1996 established national standards for health-care transactions as well as security and privacy policy for health-related information, and physicians should be aware of these standards. However, even though these regulations severely restrict the release of patient health–related information, they still permit physicians and health-care facilities to release otherwise protected health information in instances of suspected crimes or threats to public health [9,10].
Given the limited guidance for reporting suspected biocrimes, physicians could face several dilemmas. For example, some patients may not want to report a crime or a disease condition, yet may have reasonable concerns. At the beginning of the HIV/AIDS epidemic, when no treatment was available, a diagnosis of HIV infection caused many patients to fear discrimination and loss of employment, a situation that persists in many parts of the world. Similar questions may arise in bioterrorism events about insurance coverage if an event is deemed an act of war. Communication between physician and patient should help the patient understand the pros and cons of notifying law enforcement of a suspected biocrime, including whether withholding notification could place others at risk. At the very least, discussion can strengthen the doctor–patient relationship.
Fear and embarrassment of reporting potential false alarms to law enforcement or public health authorities may also be a concern for physicians and patients. But if a suspicion is not reported, a critical situation may go unrecognized and continue to worsen. Early notification to law enforcement authorities may provide valuable time and direction for investigative leads. It is expected that there will be many more negatives (false alarms) than positives when alerting law enforcement. Early reports to public health authorities may stem an epidemic. It is a misconception that you must wait for a firm diagnosis before reporting a potential case to authorities. There are many other misconceptions about biocrimes (Table 2).
Table 2 Possible Misconceptions about Biocrimes in the US
HIPAA, Health Insurance Portability and Accountability Act.
Guidance in the US concerning the reporting of suspicions of biocrimes is provided by the Centers for Disease Control and Prevention (CDC; http://www.cdc.gov), the Federal Bureau of Investigation (FBI; http://www.fbi.gov), and the Department of Homeland Security (DHS; http://www.dhs.gov) (Table 2). A joint statement by the FBI, the CDC, and the DHS advises calling the FBI and public health authorities if a suspicious situation arises [11]. Local public health departments are advised to notify the FBI before notifying the CDC. Specifically, “the FBI must be notified for any case of smallpox or pulmonary anthrax, uncommon agent or disease, an illness caused by a microorganism with markedly atypical features, an illness due to aerosol or food or water sabotage, as opposed to a usual transmission route, one or more clusters of illnesses that remain unexplained after a preliminary investigation; deliberate chemical, industrial, radiation or nuclear release” [12]. Calls or online tips should be directed to the FBI (https://tips.fbi.gov). Interpol and the World Health Organization are also developing response plans to help the public and physicians respond to suspected biocrimes and acts of bioterrorism (http://www.who.int/topics/bioterrorism/en).
The Physician's Role in Collecting Evidence
Although finding the perpetrator of a crime is a law enforcement function, the actions of attending physicians can help with microbial forensics—the scientific discipline dedicated to analyzing evidence from a biocrime or an act of bioterrorism, and that seeks to authenticate a piece of the puzzle for attribution. Implicit in the term attribution is the identification of the responsible party or the exclusion of the innocent [13].
Many physicians are familiar with the treatment of sexual assault victims, and the need to collect and preserve evidence when the patient consents. Sexual assault analysis kits have been validated to preserve semen, saliva, hair, blood, and skin. They also provide instructions on how to maintain a chain of custody to ensure that there has been no tampering with the evidence. Chain of custody is the process that assures integrity of the evidence, and ensures that there is documentation of the time the evidence is handled and each individual handling or examining the evidence. Courses exist in crime scene investigation, evidence collection, and chain of custody of the evidence in suspected sexual assault cases, and there are often well-trained support personnel who can assist patients and physicians. Such evidence collection guidance and support structures are not well-developed for biocrimes.
In contrast to typical human DNA forensic investigations, with microbial forensics, a chain of custody might not be enacted at the initial stage of medical diagnostics. Good diagnostic practices could permit samples to be used as supporting evidence in a criminal investigation. In some cases, samples can be obtained subsequently under a stricter chain-of-custody process. Law enforcement authorities can assist with such documentation processes.
Microbial Forensics
Microbial forensics includes the full scope of forensic evidence, such as analyses of microbes, materials used to prepare, stabilize, and deliver the toxin or pathogen, and fingerprints, hair, fiber, and pollen [1,14]. The laboratory analyses used for microbial forensics may include molecular sequencing, microbiological cultures, biochemistry, electron microscopy, crystallography, and mass spectrometry. These analyses go well beyond those used for medical diagnoses and epidemiologic investigations [15]. They require, however, the same substances used by the physician for diagnostics, for example, body fluid samples and microbial cultures. In this regard, the physician and the clinical laboratory have critical roles in the collection and initial analyses of samples for microbial forensics.
In the 2001 anthrax letter attacks, the preservation of the initial and subsequent isolates enabled microbial forensic methods to identify the strain in the attacks as the Ames strain of Bacillus anthracis. Analyses were based initially on a method to identify variable-number tandem repeat sequences, and later on, whole genome sequencing. Comparisons with existing strains in culture collections narrowed the likely source to a laboratory as opposed to being obtained directly from nature [16–18]. Fortunately, the initial strain from the Florida patient (the index case) and strains isolated from other victims, as well as spores from the letters, were preserved for future analyses. Microbial forensic analysis is now able, with the help of specialized facilities such as the Institute for Genome Research (TIGR), to determine the whole genome sequence of the approximately 5 million bases of B. anthracis to identify the polymorphisms that may be signatures of the bioweapon [17].
A major thrust of microbial forensics will, therefore, be the analysis of nucleic acids that can relate the genome of the pathogen to specific sources. This analysis is analogous to human DNA forensic analysis, which is being widely used to prosecute criminals and to exonerate the innocent [19]. But there are important differences between the analyses of microbial genomes and those used in human DNA forensics. Because of the sheer number of potential pathogens that could be employed as a weapon, identifying genetic markers for microbes is a more daunting task than identifying human DNA. In the case of human identification, only one species is involved, and it is often possible to identify an individual person. Viruses and most bacteria are haploid. Microbes primarily reproduce asexually, but can also evolve by recombination, horizontal gene transfer, and gene duplication. Therefore, statistical methodologies and interpretation will require different tools than are currently used for comparing and estimating the rarity of (diploid) human DNA profiles [20,21]. Nevertheless, obstacles due to genetic complexity can be reduced by obtaining samples as early as possible.
If physicians suspect a biocrime, they should take steps to ensure the preservation of the diagnostic samples so that they are not prematurely destroyed. Physicians may also advise the patient to preserve additional material that may prove useful for a criminal investigation. Just as in sexual assaults, in a suspected biocrime, the patient's personal articles may carry traditional forensic evidence that is of equal value to the information revealed by the microbe itself. Unlike sexual assault evidence, in a suspected biocrime, procedures used to preserve one particular microbe may be deleterious for other microbes and for physical evidence (such as fingerprints, culture media, isotopes, hair, and environmental material). In addition, procedures useful for preserving one microbe may be insufficient to preserve another that may be unknown at the time.
History and Physical Examination
The physician's record of the patient's history and physical examination is evidence that can be expected to be part of any public health and forensic investigation (and subsequent legal proceedings) in either a true attack or a hoax. The physician's ability to interpret the clinical history and physical examination may go beyond differential diagnoses—for example, it can help establish timelines of exposure and of the evolution of disease, which will have forensic and public health implications. The physician is positioned to assist in identification and collection of evidence, and initiate the chain of custody that protects the integrity of the evidence (see supporting online material of [1]; [14,22]) or, at a minimum, maintains good medical practice, akin to that used for transfusion of blood products.
The Case of Louisiana v. Schmidt
The case of Louisiana v. Schmidt, in which HIV-infected blood was used as the weapon in an attempted murder [23,24], is instructive for the microbial forensics system. A vial of HIV-infected blood was found in the office of a suspect, a gastroenterologist. The challenge for microbial forensics was to provide evidence that this was, or was not, the source of the victim's HIV infection. HIV, an RNA virus, undergoes rapid mutation, so any direct genetic comparison of the donor source and the recipient (the victim) is complicated. In this case, analysis focusing on both rapidly and more slowly mutating genes of HIV proved to be useful. Examination of strains from the vial, the victim, and control samples (known samples from other patients with HIV residing in the same geographic region as the victim) revealed that the viral RNA from the victim was more closely aligned to that from the vial in the suspect's office than to isolates from other patients in the area. The clinical history and clinical laboratory data obtained by attending physicians provided supporting evidence of the uninfected status of the victim prior to this injection. The victim's prior HIV-negative status was documented by blood donation screenings and negative-HIV tests of sexual partners, prior to the injection. The evidence was presented in a US criminal court. Based on the composite epidemiologic and microbial forensic evidence presented, a conviction for attempted murder was obtained.
This case illustrates several points. Sample collection and documentation by the attending physician are paramount to the biocrime investigation. If an attending physician is suspicious about an acquired infection, especially with an organism that is known to mutate rapidly, more frequent sampling and preservation of those samples are important. The sample may contain other clues (the victim, in this case, also was allegedly injected with blood that was hepatitis C positive). These samples could be helpful both epidemiologically, if there were an outbreak, and forensically, if there were an intentional incident. Analysis of these patient samples and other specimens may determine who was the source, and who was the victim. This case showed that even though the earliest isolates were not obtained, when the possibility of a biocrime was considered, there was still sufficient time to obtain valuable specimens, even with this rapidly mutating virus. The case also illustrates that microbial evidence can be informative, but it is rarely the sole deciding evidence. When considered in conjunction with other evidence—in this situation, epidemiological and clinical data—the case was very strong.
Instructive lessons can also be found by reviewing other cases in the literature [5], such as the laboratory technician who poisoned her co-workers with a laboratory stock of Shigella dysenteriae type 2 in muffins [25], and the poisoning of salad bars to skew an election for political gain [26].
Although our primary focus has been on the role that the practicing physician can play, it is important to remember that medical examiners or coroners can also serve as sentinels for discovering acts of bioterrorism and biocrime, as well as collecting pertinent microbial forensic evidence [27]. They have the statutory authority to investigate deaths that are sudden, suspicious, violent, and unattended. Moreover, the medical examiner or coroner may encounter victims that were never examined by practicing physicians. Autopsies can be crucial for diagnosis of unknown infections and for acquiring evidence for subsequent criminal investigations [1,19]. For example, in 1979, in Sverdlovsk, USSR, at least 66 people died during an anthrax outbreak. The official Soviet government position was that the victims were infected by eating contaminated meat. Autopsy data were inconsistent with the proclaimed cause of death, and, instead, supported the proposition that the disease was inhalational anthrax due to an accidental aerosol emission from a secret military weapons facility [28].
Close working relationships should be developed between the medical examiner/coroner and public health and law enforcement entities to alert one another of possible outbreaks (whether natural or intentional) as soon as possible. For suspected attacks, the medical examiner/coroner should immediately collect case-specific death investigation information and establish a chain of custody. Fortunately, medical examiners and coroners have a long-established relationship with law enforcement. However, if there are questions regarding notification or evidence collection, the medical examiner/coroner can contact the proper public health and law enforcement entities (see Table 2).
Conclusions
Physicians and other health-care providers are positioned to recognize suspicious situations and alert public health and law enforcement officials. This alone may be the most important step physicians can take (Box 2).
Box 2. Measures That a Physician May Take toward Securing Evidence in Cases of Biocrimes
Maintain primary role in caring for the patient, even at the risk of compromising evidence collection.
Discuss the situation with the patient, including options for interaction with and disclosure to public health and law enforcement officials.
If permitted by patient consent, or if required by law, alert as early as possible public health authorities and law enforcement, who can provide the necessary expertise or guidance to collect and preserve evidence.
Do not assume one agency will notify the other in a time-sensitive period. Ensure that notification has occurred.
Maintain well-documented medical records because documentation of history, physical examination, and patient course may constitute evidence.
Obtain samples that may serve as evidence early, frequently, and under a defined chain-of-custody process.
Once a biocrime is suspected, ensure that the clinical laboratory does not discard microbial isolates, but preserves them for forensic analyses or transfers them under a chain-of-custody procedure (along with accessory material such as the transport tube initially used to transport microbial isolates from patient to laboratory). Law enforcement and public health personnel participating in the Laboratory Response Network can provide this assistance.
In cases of biocrimes, physicians may interact with nonmedical authorities—who often do not fully appreciate that a trusting doctor–patient relationship is crucial for proper care and healing, and that information should be private. It is helpful to inform such officials about the importance of the doctor–patient relationship at the outset so they can be sensitive to the obligations of physicians. Just as with a sexual assault case, once there is recognition of the possibility of a bioterrorism act or other biocrime, the physician should discuss the entire situation with the patient, explaining what can be done with the consent of the patient and what actions physicians are mandated to take to comply with public health and legal requirements. This communication will likely strengthen the patient's relationship with the physician. Within the context of microbial forensics, if the patient consents, or the law requires it, the physician can facilitate preservation of evidence. To the extent possible, earlier, more, and serial sampling of evidence is best.
Physicians can ultimately serve their patients by acting, in the traditional role, as a healer, and by working with public health and law enforcement entities to help prevent further attacks and to achieve justice. As with sexual assaults, identification and conviction of the attacker can bring closure and provide a degree of security to the patient, who can then evolve from being a victim to being a survivor [29]. Physicians and their colleagues are likely to have creative ideas to contribute to the field of microbial forensics. Their input is encouraged and welcomed.
The authors are indebted to Stephen A. Morse (CDC), Barbara K. Richardson (Mt. Sinai Medical Center, New York), and Suzanne Atkin (University of Medicine and Dentistry of New Jersey—New Jersey Medical School) for their suggestions and critical review of the manuscript.
Citation: Schutzer SE, Budowle B, Atlas RM (2005) Biocrimes, microbial forensics, and the physician. PLoS Med 2(12): e337.
Abbreviations
CDCCenters for Disease Control and Prevention
DHSDepartment of Homeland Security
FBIFederal Bureau of Investigation
==== Refs
References
Budowle B Schutzer SE Einseln A Kelley LC Walsh AC Public health. Building microbial forensics as a response to bioterrorism Science 2003 301 1852 1853 14512607
Bush LM Abrams BH Beall A Johnson CC Index case of fatal inhalational anthrax due to bioterrorism in the United States N Engl J Med 2001 345 1607 1610 11704685
Marks JD Medical aspects of biologic toxins Anesthesiol Clin North America 2004 22 509 32 vii 15325716
Mayor S UK doctors warned after ricin poison found in police raid BMJ 2003 326 126
Carus WS Bioterrorism and biocrimes: The illicit use of biological agents in the 20th century center for counterproliferation research 1999 Washington (D.C.) National Defense University Available: http://www.ndu.edu/centercounter/Full_Doc.pdf . Accessed 18 August 2005
Budowle B Murch RS Chakraborty R Microbial forensics: The next forensic challenge Int J Legal Med 2005 In press
Department of Health and Human Services Select agents and toxins Title 42, Code of Federal Regulations, part 73 2003 Washington (D. C.) Department of Health and Human Services
Health Talk FDA responds to ricin baby food contamination 2004 San Francisco Health Talk Available: http://www.healthtalk.ca/ricin_07282004_8273.php . Accessed 18 August 2005
Annas GJ HIPAA regulations—A new era of medical-record privacy? N Engl J Med 2003 348 1486 1490 12686707
Hodge JG Brown EF O'Connell JP The HIPAA privacy rule and bioterrorism planning, prevention, and response Biosecur Bioterror 2004 2 73 80 15225400
Federal Bureau of Investigation, Department of Homeland Security, Centers for Disease Control and Prevention Guidance on initial responses to a suspicious letter/container with a potential biological threat 2004 Atlanta Centers for Disease Control and Prevention Available: http://www.bt.cdc.gov/planning/pdf/suspicious-package-biothreat.pdf . Accessed 18 August 2005
Centers for Disease Control and Prevention Local health officer is informed of a bioterrorist incident or threat 2001 Atlanta Centers for Disease Control and Prevention Available: http://www.bt.cdc.gov/emcontact/determine.asp . Accessed 18 August 2005
Murch RS Microbial forensics: Building a national capacity to investigate bioterrorism Biosecur Bioterror 2003 1 117 122 15040189
United States Federal Bureau of Investigation Laboratory Division Handbook of forensic services 1999 Washington (D.C.) United States Federal Bureau of Investigation Laboratory Division Available: http://www.fbi.gov/hq/lab/handbook/intro.htm Accessed 18 August 2005
Yeskey K Morse SA Roy MJ Physician recognition of bioterrorism-related diseases Physician's guide to terrorist attack 2003 Totowa Humana Press 39 46
Keim P Smith KL
Bacillus anthracis evolution and epidemiology Curr Top Microbiol Immunol 2002 271 21 32 12224522
Read TD Salzberg SL Pop M Shumway M Umayam L Comparative genome sequencing for discovery of novel polymorphisms in Bacillus anthracis
Science 2002 296 2028 2033 12004073
Pearson T Busch JD Ravel J Read TD Rhoton SD Phylogenetic discovery bias in Bacillus anthracis using single-nucleotide polymorphisms from whole-genome sequencing Proc Natl Acad Sci U S A 2004 101 13536 13541 15347815
Lander ES Budowle B DNA fingerprinting dispute laid to rest Nature 1994 371 735 738 7818670
Budowle B Genetics and attribution issues that confront the microbial forensics field Forensic Sci Int 2004 146 S185 S188 15639572
Budowle B Chakraborty R Doutremepuich C Morling N Genetic considerations for interpreting molecular microbial forensic evidence Progress in forensic genetics 10 2004 Amsterdam Elsevier 56 58
American Society of Crime Laboratory Directors/Laboratory Accreditation Board Laboratory management and operations 2004 Garner (North Carolina) American Society of Crime Laboratory Directors/Laboratory Accreditation Board Available: http://www.ascld-lab.org/legacy/aslablegacymanagement.html . Accessed 18 August 2005
Metzker ML Mindell DP Liu XM Ptak RG Gibbs RA Molecular evidence of HIV-1 transmission in a criminal case Proc Natl Acad Sci U S A 2002 99 14292 14297 12388776
Heitpas J McMullen LK Mindell DP Hanson HL Rice CM Breeze RG Budowle B Schutzer SE Keeping track of viruses Microbial forensics 2005 San Diego Academic Press 55 97
Kolavic SA Kimura A Simons SL Slutsker L Barth S An outbreak of Shigella dysenteriae type 2 among laboratory workers due to intentional food contamination JAMA 1997 278 396 398 9244331
Torok TJ Tauxe RV Wise RP Livengood JR Sokolow R A large community outbreak of salmonellosis caused by intentional contamination of restaurant salad bars JAMA 1997 278 389 395 9244330
Nolte KD Hanzlick RL Payne DC Kroger AT Oliver WR Medical examiners, coroners, and biologic terrorism: A guidebook for surveillance and case management MMWR Recomm Rep 2004 53 1 27
Jackson PJ Hugh-Jones ME Adair DM Green G Hill KK PCR analysis of tissue samples from the 1979 Sverdlovsk anthrax victims: The presence of multiple Bacillus anthracis strains in different victims Proc Natl Acad Sci U S A 1998 95 1224 1229 9448313
Hampton HL Care of the woman who has been raped N Engl J Med 1995 332 234 237 7619112
Breeze R Budowie B Schutzer S Microbial Forensics 2005 San Diego Academic Press 448
|
16167845
|
PMC1236212
|
CC BY
|
2021-01-05 10:39:22
|
no
|
PLoS Med. 2005 Dec 27; 2(12):e337
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020337
|
oa_comm
|
==== Front
Mol CancerMolecular Cancer1476-4598BioMed Central London 1476-4598-4-261604275910.1186/1476-4598-4-26ResearchVariation in gene expression patterns in effusions and primary tumors from serous ovarian cancer patients Schaner Marci E [email protected] Ben [email protected] Martina [email protected] Reuven [email protected]ørenes Vivi Ann [email protected] Björn [email protected] Aasmund [email protected] Iris [email protected] Vered [email protected]è Claes G [email protected] Gunnar B [email protected] Jahn M [email protected]ørresen-Dale Anne-Lise [email protected] Departments of Biochemistry (M.E.S.), Stanford University School of Medicine, Stanford, CA 94305-5151, USA2 Department of Pathology, The Norwegian Radium Hospital, Montebello N-0310 Oslo, University of Oslo, Norway3 Department of Pharmacology and Experimental Therapeutics, Faculty of Medicine, Hebrew University, Jerusalem 91120, Israel4 Department of Gynecologic Oncology, The Norwegian Radium Hospital, University of Oslo, Montebello N-0310 Oslo, Norway5 Department of Genetics, The Norwegian Radium Hospital, University of Oslo, Montebello N-0310 Oslo, Norway6 Deceased2005 21 7 2005 4 26 26 22 4 2005 21 7 2005 Copyright © 2005 Schaner et al; licensee BioMed Central Ltd.2005Schaner 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 numerous studies have characterized primary ovarian tumors, little information is available regarding expression patterns of metastatic sites of this cancer. To define sets of genes that distinguish primary and metastatic ovarian tumors, we used cDNA microarrays to characterize global gene expression patterns in 38 effusions (28 peritoneal, 10 pleural) and 8 corresponding primary ovarian tumors, and searched for associations between expression patterns and clinical parameters.
Results
We observed multidimensional variation in expression patterns among the cancers. Coordinate variation in expression of genes from two chromosomal regions, 8q and 19q, was seen in subsets of the cancers indicating possible amplifications in these regions. A set of 112 unique genes of known function was differentially expressed between primary tumors and effusions using supervised analysis. Relatively few differences were seen between effusions isolated from the pleural and peritoneal cavities or between effusions from patients diagnosed with stage III and stage IV cancers. A set of 84 unique genes was identified that distinguished high from lower grade ovarian cancers. The results were corroborated using immunocytochemistry, mRNA in situ hybridization, and immunoblotting.
Conclusion
The extensive variation in expression patterns observed underscores the molecular heterogeneity of ovarian cancer, but suggests a similar molecular profile for ovarian carcinoma cells in serosal cavities.
==== Body
Background
Epithelial ovarian carcinoma claims more lives than any other gynecologic malignancy, largely because it frequently escapes detection after it has metastasized [1]. Ovarian carcinoma initially metastasizes primarily to the serosal surface of the peritoneal cavity and abdominal organs. The pleural space is often involved as well, either at diagnosis or, more commonly, at later stages of clinical progression. Pleural effusion is the most common presentation of stage IV disease [2]. A number of metastasis-associated molecules have been reported to be differentially expressed between primary ovarian tumors and tumor cells in effusions [3-12], but little is known regarding the mechanism of metastases.
Molecular characterization of ovarian carcinoma using DNA microarrays has so far focused on primary tumors [13-22]. The paucity of data regarding the biological characteristics of ovarian carcinoma cells in effusions at both the phenotypic and genotypic level limits our understanding of tumor progression in this disease. Specifically, we do not know how ovarian carcinoma cells in ascites and pleural effusions differ from those in the corresponding solid primary tumors, or whether and how carcinoma cells in peritoneal and pleural effusions differ. Moreover, molecular analysis of malignant effusions might contribute to better predictions of survival and treatment response.
To identify genes whose expression may be associated with this metastatic behavior, we analyzed global gene expression patterns of ovarian cancer cells obtained from 3 distinctive anatomic sites: 28 peritoneal, 10 pleural and 8 primary tumors (see supplementary Table S1.xls). A valuable feature of this dataset is that it includes 8 paired samples of primary tumors and malignant effusions from the same patients. We were able to define a number of genes that differentiate primary tumors from effusions.
Results
Overview of global gene expression patterns among ovarian cancers
We profiled 46 ovarian tumor samples, 38 effusions and 8 primary ovarian carcinomas (Figure 1A–C) using cDNA arrays representing approximately 26,965 genes and selected those genes that passed a simple data quality and variation filter (see Materials and Methods). Using hierarchical clustering of the 2863 genes that passed our filtering criteria, we found considerable heterogeneity in the expression patterns among the tumor samples. The clustering analyses divided the ovarian cancer specimens into two major groups, with 4 of the 8 primary tumors clustering together but apart from their paired effusions. It is noteworthy that the other 4 primaries clustered together with the effusions from the same patient (Figure 1B). The major distinguishing feature between the two branches of the dendrogram was high expression of a number of chemokines, collagens, cell surface antigens, adhesion molecules and leukocyte antigens (Figure 1A, panels g, h). Some of the cancers were notable for the elevated expression of a cluster of genes residing on chromosome segment 8q21-24 and the coordinate variation in expression of these genes suggests that there may be an amplification of this region of chromosome 8 in some of the ovarian cancers (Figures 1C, panel b highlighted in red. See also Figure 4C, panel a). The cancers with chromosome 8q21-24 overexpression were mostly the paired primary tumors and effusions.
Figure 1 Overview of Primary Tumors and Effusions. (A) Global gene expression patterns of 46 ovarian cancers: 8 primary tumors, 10 pleural effusions and 28 peritoneal effusions, were sorted based on similarity of expression following hierarchical clustering. 2863 genes were selected from the total data set based on variance greater than 2.5 fold in at least 3 arrays. (B) The dendrogram is color-coded pink (pleural effusions), blue (peritoneal effusions) and black (primary tumors) to indicate site of origin of the cancers. Indicates the clear cell sample. (C) Magnified view of specific gene clusters selected from the entire set of 2863 genes: (a) Caveolin 1 and EGFR, (b) Chromosome 8 (Genes on Chromosome 8 are denoted in red), (c) Cell Cycle Associated Genes, (d) Kallikreins 6, 10, (e) Epithelial, (f) Sprouty cluster, (g) Stromal cluster, (h) immune response cluster. The scale is indicated in the bottom right-hand corner and spans 0.25 to 4 fold over mean (-2 to +2 in log2 space). Missing data are denoted in gray.
Figure 4 Overview of PAM results following clustering. PAM was carried out to determine differences between the 3 sites examined in this study: Primary tumors (Black), peritoneal effusions (Blue) and pleural effusions (Pink). Three main clusters differentiate the groups: (a) a set of genes that divides the cluster in to 2 groups and is more highly expressed in the majority of ascites (genes mapping to chromosome 8 are denoted in red), (b) a set of genes more highly expressed among the pleural effusions, but also expressed in a subset of the ascites, (c) Large cluster over-expressed among the primary tumors.
Genes involved in cell cycle progression and cell proliferation were variably expressed among the cancers presumably reflecting variation in proliferation rates among the tumors in a coordinated manner (Figure 1C, panel c) [23,24]. Co-expression of previously identified markers for ovarian cancer including kallikreins 6 and 10 as well as the S100 calcium-binding proteins S100A1 and S100A13 (Figure 1C, panel d) was also seen. Co-clustering of CEACAM5 and CEACAM7 with syndecan-1 (Figure 1C, panel e) was observed. These genes were part of a larger cluster with expression of Keratins 5, 6B, 7, 15 and 19, and S100A10, caveolin-2 and SLPI (Figure 1C, panel e). In another cluster (Figure 1C, panel f) the sprouty homologs: 1, 2, and 4, were co-expressed with the dual specificity phosphatase 6 (DUSP 6) and two ets variant genes 4 and 5 (ETV4 and ETV5), indicating possible involvement of the MAP kinase pathway in some tumors. Also, co-expression of caveolin 1 and EGFR was seen, in agreement with data from experimental models [25,26] (Figure 1C, panel a). The complete gene list and clustered file is found in supplementary Figure S1 CDT.cdt.
In order to examine the concordance between expression at the mRNA level and protein expression, we carried out immunocytochemistry of CD44, ITGB3, and CD168 (Syndecan-1) on selected samples and show representative stains (Figure 2, panels A-C). Comparing the ICC results with the expression result gave good correlation. Of the 13 samples having both expression data and ICC for CD44, the 4 with positive ICC had a log2 ratio of expression from 1 to 2, and the 9 ICC negative had log2 ratio from 0 to -2. For the ITGB3 11 samples had both ICC and expression data, and of the 3 ICC positive the log2 ratios were from 1 to 2. Of the 8 ICC negatives 7 had expression ratios from 0 to -2 and one did not correspond with the expression level with a log2 ratio 1.5. For the Syndecan 1 (CD 168) of the 9 ICC positive samples 4 had log2 ratios from 1.5 to 2, 4 had log2 ratios of 0 and 1 had a log2 value of -0.5. The 3 ICC negatives had expression ratios from -1 to -2.
Figure 2 Immunocytochemistry (ICC) of selected proteins. ICC of (A) CD44, (B) CD61, and (C) Syndecan-1 (CD168) was carried out as detailed in 'Materials and Methods'. Expression Data is shown in the top panel and corresponding cases with protein staining are denoted with arrows. In panel A the following samples are shown: OV 06A and OV 34P. In panel B the following samples are shown: OV 02A, OV12A and OV14A. In panel C the following samples are shown: OV 02A, OV18A and OV20A.
In addition, localization to the tumor cells was confirmed. In situ hybridization was carried out for the ets variant ETV4 (PEA3) (23 cases) as well as MMP-9 (19 cases) and MMP-14 (13 cases), demonstrating high mRNA expression in the tumor cells (Figure 3). These molecules were chosen since they have been shown to be involved in the metastatic process and are expressed in ovarian carcinomas and to be of prognostic value. Negative and positive controls for ICC and ISH showed consistent results through all experiments.
Figure 3 In situ hybridization of selected genes. mRNA in situ hybridization in ovarian carcinoma effusions. Negative control specimen (stained with nuclear fast red) and d(T) control are shown in the first row. Two negative (left, stained red) and two positive (right, gray) cases using the PEA3 probe are shown in the second row. Two positive (left, stained gray) and one negative (right, red) cases using the MMP-9 probe are shown in the third row. Three positive cases using the MMP14 probe are shown in the fourth row.
Expression differences between primary tumors, peritoneal effusions and pleural effusions
To capture any differences in expression profiles between primary tumors versus all effusions, SAM analysis was performed on 2863 genes that passed the described filter criteria (see Materials and Methods). By this analysis a set of 112 unique genes of known function that were differentially expressed in these two groups was identified. (False discovery rate 2%). A partial list of the unique genes is shown in Table 1 and the complete list may be found in supplementary Table S2.xls. The relative levels of expression of several epithelial markers, including claudin 7 and keratins 7 and 19 were higher in the effusions. The expression of some genes characteristically expressed in stromal cells, including SPARC and collagens 1A1, 5A2, and 6A2 was generally higher in the primary tumors than in the effusions.
Table 1 Selected genes identified by SAM that are differentially expressed in primary tumors vs. all effusions. The complete list is available in Table S2.
Genes more highly expressed in Effusions
CLDN7 : claudin 7 : Hs.278562
KRT7 : keratin 7 : Hs.23881
CRIP1 : cysteine-rich protein 1 (intestinal) : Hs.423190
L1CAM : L1 cell adhesion molecule MASA: Hs.1757
ADM : adrenomedullin : Hs.394
CRYAB : crystallin, alpha B : Hs.391270
IL18RAP : interleukin 18 receptor accessory protein : Hs.158315
KRT19 : keratin 19 : Hs.182265
Genes more highly expressed in Primary Tumors
COL6A3 : collagen, type VI, alpha 3 : Hs.80988
COL1A2 : collagen, type I, alpha 2 : Hs.179573
MYB : v-myb myeloblastosis viral oncogene homolog (avian) : Hs.1334
BGN : biglycan : Hs.821
IGFBP7 : insulin-like growth factor binding protein 7 : Hs.119206
SPARC : secreted protein, acidic, cysteine-rich (osteonectin) : Hs.111779
AEBP1 : AE binding protein 1 : Hs.118397
APOD : apolipoprotein D : Hs.75736
CDH11 : cadherin 11, type 2, OB-cadherin (osteoblast) : Hs.75929
The filtered data for the effusions only were then used to examine gene expression differences among pleural and peritoneal effusions. Using SAM only 19 unique genes of known function were identified that significantly varied between pleural and peritoneal effusions (Table 2), with a false discovery rate of 13.5%. Pleural effusions had significantly higher expression of the angiogenic inducer CYR61, RAB21, glutathione S-transferase A4 (GSTA4), and several chemokines when compared to peritoneal effusions. In addition, expression of the iron transporter, SLC40A1 was generally lower in pleural effusions than in the ascites samples. Although some differences in expression were seen between pleural and peritoneal effusions, the results provide evidence in favor of a more similar genetic profile for cancer cells at these two anatomic sites than for effusions versus the primary tumor.
Table 2 Genes identified by SAM that are differentially expressed in Pleural vs. Peritoneal Effusions
Positive Significant Genes: Higher in pleural vs. ascites
NR4A1: nuclear receptor subfamily 4, group A, member 1: Hs.1119
CYR61: cysteine-rich, angiogenic inducer, 61: Hs.8867
CXCL2: chemokine (C-X-C motif) ligand 2: Hs.75765
RAB21: member RAS oncogene family: Hs.184627
CTGF: connective tissue growth factor: Hs.75511
CXCL3: chemokine (C-X-C motif) ligand 3: Hs.89690
TCEB3: transcription elongation factor B (SIII): Hs.155202
IGLL1: immunoglobulin lambda-like polypeptide 1: Hs.348935
CTGF: connective tissue growth factor: Hs.75511
IGHG3: immunoglobulin heavy constant gamma 3: Hs.413826
C1QB: complement component 1, q subcomponent beta: Hs.8986
CYR61: cysteine-rich, angiogenic inducer, 61: Hs.8867
C1QG: complement component 1, q subcomponent gamma:Hs.94953
TAGLN: transgelin: Hs.433399
CD163: Hs.74076
GSTA4: glutathione S-transferase A4: Hs.169907
Negative Significant Genes: Lower in pleural vs. ascites
N33: Putative prostate cancer tumor suppressor: Hs.71119 100791
PLEC1: plectin 1, intermediate filament binding protein 500 kDa: Hs.79706
SLC40A1: solute carrier family 40, member 1: Hs.5944
We then applied a supervised statistical method, PAM [27] to see whether it was possible to find a set of genes that could classify the primary tumors from peritoneal and pleural effusions. By comparing the three groups, a set of genes (615; 436 unique genes) was identified that correctly classified the primary tumors and most of the peritoneal (error rate 10%) and pleural effusions (error rate 20%). Cross-validation was not as successful, suggesting again that the difference between the 2 groups of effusions is not as clear as the distinction between all effusions and the primary tumors, which is in accordance to the results of the SAM analyses. Some differences were however detected. Clustering of the 615 genes (Figure 4 and supplementary Tables S3a, S3b.xls) illustrates that the expression of a number of chemokine ligands, including CCL2, CCL8 and CCL3L1 was more frequent in pleural effusions possibly reflecting the larger number of leukocytes in these specimens (Figure 4C, panel b). Higher expression of a number of genes on Chromosome 8q24 was observed among most of the ascites (Figure 4C, panel a highlighted in red). In addition, the expression of genes whose proteins have previously been shown to be produced by both ovarian carcinoma and stromal cells [28-30], including TIMP-2, vimentin and basic fibroblast growth factor separated the primary tumors from the effusions, suggesting that the cancer-stroma crosstalk is associated with different biological pathway activation than that observed in effusions (Figure 4C, panel c and supplementary Table S2.xls).
Hierarchical clustering of all 38 effusions was carried out as described previously. Results were very similar to those obtained when clustering both the primary tumors and effusions (Web Supplement, Figure S2.pdf) with the same 11 samples residing on the left branch as in Fig. 1 and with a strong overexpression of the immune response cluster of genes (Figure 1h). Interestingly, the majority of the pleural effusions (6/10) clustered to this branch (p = 0.02).
Identification of Expression Profiles Based on 'Intrinsic' Genelist
An 'intrinsic' genelist was constructed to further analyze the data (see Materials and Methods). The main clusters identified using this genelist highlighted probable regions of chromosomal changes on chromosomes 8 and 19 in a subset of tumors (Figure 5C, panels a, b). The majority of genes in each respective cluster mapped to either 8q21-24 or 19q13. In this analysis, all effusion-primary tumor pairs clustered together as expected. Furthermore, multi-dimensional variation was notable in groups of kallikreins 6 and 10, S100A1, S100A13, EPAC, laminin γ2 (LAMC2), MUC5B, TRIM 29 and claudin 10 (Figure 5C, panel c). The kallikreins have been shown to be potentially useful prognostic markers in ovarian cancer [31,32] and laminin γ2 and MUC5B were shown to display high expression among some of the cancers at the protein level as well, using Western analyses (Figure 6). Correlation between the protein expression and mRNA expression showed that 3 of the 4 effusions positive for MUC5B protein had positive log2 ratios. All the cases lacking protein expression had negative log2 ratios. For the laminin γ2 chain, the protein expression level did not correspond that well with the mRNA expression although all the protein-negatives had negative log2 ratios and the majority of the strongly positive effusion had positive mRNA expression.
Figure 5 'Intrinsic' cluster: Overview of hierarchical clustering with the 'Intrinsic' genelist Overview of 2121 genes selected and magnified view of specific clusters. (a) Chromosome 8 associated cluster, (b) Chromosome 19 associated cluster, (c) Variation in expression of kallikreins and other genes among the cancers.
Figure 6 Western blotting of selected proteins. Immunoblotting of 18 effusions using antibodies directed against Mucin B5 and the laminin γ-chain (Santa Cruz). Mucin B5 (upper panel) is expressed in the specimens in lanes 1,2,3,5,12 and 14. The laminin γ-chain is expressed in all specimens except those in lanes 5 and 13. The specimens analyzed were from1 to 17: OV08A, OV09A, OV10A, OV12A, OV11A, OV02A, OV04A, OV29P, OV06A, OV01A, OV32P, OV14A, OV12A2, OV18A, OV20A, Additional specimen not on array, OV37P.
The relationship between FIGO stage and gene expression patterns was examined using the genes identified in the intrinsic list. Using SAM, we identified only 7 named genes and one undefined clone (False Positives, 2.9) that were more highly expressed in effusions from late stage (Stage IV) ovarian cancer. Among these genes were PEN2 (presenilin enhancer 2) and PDCD5 (programmed cell death 5) both residing on 19 q12-13, (Web Supplement, Table S4.xls). Since most of the stage IV cases in this study were defined as such by the presence of pleural effusion, this further underscores the few differences between cells in effusions in Stage IIIc and Stage IV disease. We identified 84 genes that distinguished low from high grade disease (False Discovery Rate of 12.9%, Table 3, Web Supplement, Table S5.xls) including MAGE A6, a member of a family of proteins that may be useful at selectively distinguishing cancer cells from normal cells that do not express this antigen. These genes may be important in understanding the progression of ovarian cancer.
Table 3 representative list is shown below and the complete list is available in Table S5.
DERP6: S-phase 2 protein: Hs.417029
VARS2L: valyl-tRNA synthetase 2-like: Hs.102910
SC5DL: sterol-C5-desaturase (ERG3 delta-5-desaturase homolog, fungal)-like: Hs.287749
MYLIP: myosin regulatory light chain interacting protein: Hs.443793
HMGN3: high mobility group nucleosomal binding domain 3: Hs.77558
MAGEA6: melanoma antigen, family A, 6: Hs.441113
ATP5L2: ATP synthase, H+ transporting, mitochondrial F0 complex, subunit g, isoform 2
GPR48: G protein-coupled receptor 48: Hs.160271
MASK: multiple ankyrin repeats, single KH-domain (MASK) homolog: Hs.528646
PRO1853: hypothetical protein PRO1853: Hs.433466
YKT6: SNARE protein Ykt6: Hs.296244
ZNF-kaiso: kaiso: Hs.143604
NCK1: NCK adaptor protein 1: Hs.54589
Discussion
We examined global gene expression patterns in primary ovarian cancers and malignant effusions and found differences that may be related to tumor progression. Our results are consistent with many reports that document the heterogeneity of gene expression patterns noted in ovarian cancer [14,16-18,20-22]. These data underscore the heterogeneity of this disease and the profound molecular differences within tumor sub-groups with comparable morphology.
Several clusters of variably expressed genes may have relevance to the biology of ovarian cancer. A large cluster of genes from chromosome 8q21-24 was more highly expressed among a subset of the cancers, suggesting an amplification of this region on chromosome 8. This was seen in both the primary and the effusions from the corresponding samples with these abnormalities. Transcripts over-expressed in this cluster include YWHAZ, that encodes the zeta isoform of 14-3-3 protein (tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein) family. 14-3-3 proteins are expressed in a number of cancers and are involved in the cell cycle and also in prolonging cell survival. TPD52 (hD52) that has been shown to be in the peak of the 8q21 amplicon in breast cancer cell lines was also over-expressed in a number of cancers in this cluster. A recent report suggests that TPD52 is a candidate target gene and putative oncogene at 8q21 [33]. In addition, co-clustering of a number of genes in this region was also observed in the cluster resulting from the intrinsic genelist (Figure 5C, panel a).
Involvement of the MAP kinase pathway may be indicated in a large number of the cancers in our study based on the co-expression of the 3 sprouty transcripts (1, 2, and 4) (Figure 1C panel f). Expression of all three sprouty homologs has been observed before in mouse development [34], but not in human cancers. Human sprouty 2 has been shown to inhibit the mitogen-activated protein (MAP) kinase pathway [35], and sprouty proteins have been implicated in the negative regulation of the receptor tyrosine kinase-induced MAP kinase pathway. DUSP, another gene in our sprouty cluster, belongs to a family of 9 cytoplasmic and/or nuclear enzymes, that function by dephosphorylating threonine and tyrosine residues of p38, JNK and ERK [36,37]. The transcriptional regulation of Ets transcription factors, including that of PEA3 (ETV4), on the other hand, involves signals initiated by growth factor signaling through tyrosine kinase receptors that may be mediated by MAPK [38]. We have previously shown that PEA3 and DUSP member PAC-1 predict poor outcome in ovarian carcinoma, whereas expression of all three MAPK predicts improved survival [39-41]. The coordinated expression of 3 sprouty family members, DUSP6, and ETV4 and 5 in a subset of ovarian cancers therefore raises the possibility that the activity of a specific, MAPK-related signaling pathway may have a role in ovarian cancer.
We have previously shown that clinical and molecular markers that are of established prognostic role in primary ovarian cancer have little or no significance in effusions [42]. Besides the obvious fact that cells at this site represent tumor progression, they also appear to be biologically different than tumor cells in both primary tumors and solid metastases. Using PAM to analyze the three different groups, primary tumors, and pleural and peritoneal effusions, we identified a subset of genes that is more highly expressed in the primary tumors, but fewer differences between pleural and peritoneal effusions. These genes are mainly characteristic of previously defined 'stromal' signature [43], but that are produced by both tumor and stromal cells including collagens, TIMP2, bFGF, vimentin and SPARC [28-30]. To study tumor progression we evaluated the expression levels in relation to clinical parameters using SAM, focusing on FIGO stage and grade. The most notable distinctions were based on grade rather than stage.
A large cluster of genes on chromosome segment 19q13.1 showed covarying expression among the cancers in this study, consistent with previous reports that the 19q13.1 region is amplified in some ovarian carcinomas [44,45]. The AKT2 oncogene contained in this region of chromosome 19, has been shown to be associated with the progression of ovarian cancer [44,45]. Interestingly, we found that some of the genes in the 'chromosome 19 cluster' including PDCD5 and PEN2, were more highly expressed in the Stage IV cancers than earlier-stage supporting the notion that amplification of the 19q13.1 region and concomitant elevated expression of these genes may play a role in the progression of the disease [46]. Furthermore in a larger cohort of advanced ovarian carcinomas amplification of this region was found to be associated with poor survival (Wang et. al., manuscript to be submitted).
In summary, we have examined the relationship between primary tumors in ovarian cancer and their corresponding effusions. The most notable difference was observed in expression patterns between effusions and the primary tumors. There is significant molecular variation among the cancers. There are some hints of differences related to tumor grade and effusion versus primary tumor that will need further investigation to see whether these are significant. Finally, some consistent features of expression patterns in subsets of the cancers may suggest possible molecular alterations involved in the biology of the tumors.
Materials and methods
Effusion specimens
Material consisted of 38 fresh non-fixed peritoneal and pleural effusions submitted from the Department of Gynecological Oncology to the Division of Cytology, Department of Pathology, The Norwegian Radium Hospital, during the period of April 1998-May 2002. Specimens were obtained pre-operatively, intra-operatively, or at disease recurrence, from 36 patients diagnosed with ovarian carcinoma (35 of the serous type, one clear cell type) and one patient diagnosed with primary peritoneal carcinoma (PPC). From one patient, two ascites specimens were obtained one month apart. Effusion specimens consisted of 28 peritoneal and 10 pleural effusions. Patient age ranged from 35 to 73 years (mean = 58 years). Twenty patients were diagnosed with FIGO stage III disease and 16 with stage IV disease. The remaining patient had a stage I tumor. Tumor grade for 36 serous carcinomas was as follows: 5 grade 1, 10 grade 2 and 21 grade 3 carcinomas. All relevant clinical data were obtained from the Department of Gynecologic Oncology, and are presented in detail in web supplementary Table S1.xls. In order to preserve physiological activity, specimens submitted to our laboratory arrived within minutes after collection and were processed immediately. Cells were suspended and frozen in RPMI+DMSO at -70°C. Smears and cellblock sections from all specimens underwent morphological evaluation by three experienced cytopathologists, and were further characterized using immunocytochemistry with broad antibody panels against epithelial and mesothelial epitopes, as previously detailed [5,6,47]. In all specimens included in this study, cancer cells comprised 50% or more of the entire cell population based on cytology smears. All patients in this study were treated with platinum based chemotherapy according to current guidelines, and the samples were collected under an IRB approved protocol (S-01127; June 22, 2001).
Solid tumors
We obtained samples of primary tumors from 8 of the patients whose effusions were analyzed. These were snap-frozen in liquid nitrogen upon removal and stored at -70°C. Frozen sections were obtained from all biopsies in order to evaluate the percentage of tumor cells and tissue viability. The former exceeded 50% of cells in all cases.
RNA isolation, Labeling and Hybridization
Total RNA was isolated from effusions and solid tumors using the TRIZOL Reagent (Gibco BRL, Life Technologies). mRNA isolation from total RNA was undertaken using d(T) coated Dynabeads (Dynal, Oslo, Norway). One to two μg of experimental sample mRNA was used for labeling with Cy5-dUTP. mRNA was reverse transcribed with Superscript II (Invitrogen Life Technologies) using an oligo dT primer (Operon Technologies, Alameda, CA). Each sample was comparatively hybridized to cDNA microarrays together with a common reference labeled with Cy3-dUTP (Stratagene). Fluorescent dyes were purchased from Amersham Pharmacia Biotech, Piscataway, NJ. Hybridizations were carried out using the standard protocol described previously. [23,24]. Complete experimental details may be found at: .
DNA Microarrays: All arrays were printed at the Stanford Functional Genomics Facility. DNA clones on the microarrays are based primarily the sequence verified IMAGE clones from the Research Genetics Corporation (Huntsville, AL) and the CGAP clone set as well as a small percentage of custom spots and control spots. Complete details regarding the clones on the arrays may be found at . These microarrays were comprised of 41,805 elements (42 K), representing an estimated 25,695 genes as judged by the number of unique Unigene symbols.
Data Analysis and Clustering
Data Selection: Data were analyzed by using either the GenePix 3.0 or GenePix 4.0 software (Axon Instruments). Spots with aberrant measurements due to obvious array artifacts or poor technical quality were manually flagged and removed from further analysis. A filter was applied to omit measurements where fluorescent signal from the DNA spot was less than 50% above the measured background fluorescence surrounding the printed DNA spot in both the Cy3 and Cy5 Channels. Genes that did not meet these criteria for at least 80% of the measurements across the cases were excluded from further analysis. Data were retrieved as log2(Cy5/Cy3). The (Cy5/Cy3) ratio is defined in Stanford Microarray Database (SMD) as the normalized ratio of the background-corrected intensities [48]. Genes whose expression level differed by more than 2.5-fold from their mean expression level in the sample set in at least 3 samples, were selected for further analysis.
Significance Analysis of Microarrays (SAM)
SAM is a statistical approach to identify genes whose expression patterns are significantly associated with specific characteristics of sample sets [49] SAM analysis was applied to the ovarian dataset to examine differences between primary, pleural and peritoneal effusions, and to examine different clinical parameters, including stage and grade. A two-way, unpaired test was carried out comparing the two groups of interest. A 10-nearest neighbor imputation engine was applied to estimate missing data [50], and 500 permutations were carried out to compute expected values and to calibrate false positive calls.
Prediction Analysis for Microarrays (PAM)
Prediction Analysis for Microarrays (PAM) was carried out using the Excel version of the program. This method is applied to gene expression data to provide sample classification by 'shrunken centroids'. Data were filtered as described earlier for clustering analyses. The three groups used for analysis were defined as primary tumors, peritoneal effusions (ascites) and pleural effusions [27].
Intrinsic Genelist
An 'intrinsic' genelist, comprising genes whose expression varied significantly more between samples from different patients than between replicate samples for the same patient, was selected based on the 8 primary tumor/effusion pairs and the paired effusion cancers using methods as described previously [23]. A score was constructed which was the average within-pair variation relative to the between-pair variation. This was the ratio of the variance of the differences, relative to the variance of the averages.
Immunocytochemical analysis (ICC)
ICC was performed using antibodies directed against Syndecan-1 (CD138; SDC1), the integrin β3 subunit (platelet glycoprotein IIIa, antigen CD61), and CD44 antigen, all part of identified gene clusters showing variable degrees of expression among the samples. Negative controls consisted of sections that underwent a similar staining procedure, with the exclusion of primary antibody application, or that were stained with mouse myeloma protein of the same isotype as the primary antibody used. Biopsies in which immunoreactivity for the studied antigens had previously been demonstrated were used as positive controls.
Western blotting
Frozen effusion specimens were thawed and washed twice in phosphate buffered saline (PBS). Samples were subsequently lysed in 1% NP-40, 20 mM Tris HCl (pH 7.5), 137 mM NaCl, 10% glycerol, 1 mM Phenylmethylsulfonyl-fluoride, and 1 mM Sodium Orthovanadate, with 0.02 mg/ml aprotinin, leupeptin and pepstatin and 10 μg/ml phosphatase inhibitor cocktail I. All inhibitors were from Sigma-Aldrich (Saint Louis, MO). After centrifugation, the supernatant was collected and protein content was evaluated by the Bradford assay. Twenty-five μg total protein lysate/lane was resolved by SDS polyacrylamide gel electrophoresis (7.5 or 12 % gels) and transferred on to PVDF immobilon membranes (Millipore, Bedford, MA). Successful transfer was evaluated by staining of membranes withNaphtol-blue-black (Sigma-Aldrich). Filters were blocked in TBST containing 5 % dried skimmed milk and 0.05% Tween-20 for 1 hr at room temperature. Thereafter, the filters were incubated over night at 4°C with primary antibodies diluted in TBST containing 5 % skimmed milk and 0.05% Tween-20. Primary antibodies directed against the laminin γ-2 chain (Santa Cruz Biotechnology, Santa Cruz, CA) and Muc 5B (Santa Cruz) were chosen as examples since they were part of novel clusters displaying variation in expression among the cancers. Filters were washed 3 times 10 minutes each with TBST (0.05% Tween-20). They were subsequently incubated with HRP-conjugated secondary antibody diluted 1:5000 in TBST containing 5% dried skimmed milk and 0.05% Tween-20 for 45 minutes at room temperature. Immunoreactivity was detected using the ECL-plus western blotting system (Amersham-Pharmacia)
mRNA In Situ hybridization (ISH)
Three genes were chosen for ISH as representative examples for the evaluation of mRNA expression levels in the tumor cells themselves: MMP-14, MMP-9, and PEA3 (ETV4). The following antisense oligonucleotide probes were obtained from Research Genetics (Huntsville, AL) [1-28,51,52]:
MMP-9: 5' CCGGTCCACCTCGCTGGCGCTCCGGU 3', PEA3: 5' TGA ATT ATG AGA AGC TGA GCC G 3', MMP-14: 5' TCC ATC ACT TGG TTA TTC CTC ACC CGC C 3'.
A poly d(T)20 oligonucleotide (Research Genetics) was used to verify the integrity and lack of degradation of mRNA in each sample. The DNA probes were hyperbiotinylated. Stock dilution was prepared with a resulting equal concentration for both probes. The stock dilution was diluted with probe diluent (Research Genetics) immediately before use. Specific sense oligonucleotides were used for the evaluation of non-specific activity for each probe.
Cellblock sections (4 micron-thick) of formalin-fixed, paraffin-embedded specimens were mounted on ProbeOn Plus slides (Fisher Scientific, Pittsburgh, PA). Sectioning was performed in RNase-free water. Hybridization using the probes was carried out as previously described and by using the microprobe manual staining system (Fisher Scientific)[53]. Known positive controls were used in each hybridization reaction. These consisted of 2 ovarian carcinomas for which positive hybridization was reproducible in a previous study. Controls for endogenous alkaline phosphatase for all probes included treatment of the sample in the absence of the probe and use of chromogen alone.
Authors' contributions
ALBD, MES, BD designed the experiments and wrote the manuscript. CGT, JBK, JMN, AB, BR, BD, collected the samples and collected and analyzed the clinical data, MES, BD, RR, VAF, IG, VG, MS performed experiments.
Supplementary Material
Additional File 1
Figure S1: CDT file for fig 1
Click here for file
Additional File 2
Figure S2: Effusion only Cluster
Click here for file
Additional File 3
Table S1: Full Clinical Data
Click here for file
Additional File 4
Table S2: SAM 153 full list primary tumors vs. effusions
Click here for file
Additional File 5
Tables S3a, S3b: Full list of the PAM analysis comparing the primary tumors and effusions
Click here for file
Additional File 6
Table S4: Analysis of stage using SAM
Click here for file
Additional File 7
Table S5: Analysis of grade using SAM (full list)
Click here for file
Acknowledgements
This work was supported by grants from the Norwegian Cancer Society (D 99061), The Research Council of Norway (155218/300), NIH Grant 2HFZ542 and the Marsha Rivkin Scholar Award (M.E.S.). We wish to thank Mike Fero, and the staff of the Stanford Functional Genomics Facility, members of the Brown lab for helpful discussions, and the Stanford Microarray Database, with special thanks to Jeremy Gollub and Gavin Sherlock. Trevor Hastie provided advice and constructed the intrinsic genelist. We also would like to thank Pat Brown for helpful discussions, comments and support.
==== Refs
Greenlee RT Hill-Harmon MB Murray T Thun M Cancer statistics, CA Cancer J Clin 2001 51 15 36 11577478
Akahira JI Yoshikawa H Shimizu Y Tsunematsu R Hirakawa T Kuramoto H Shiromizu K Kuzuya K Kamura T Kikuchi Y Kodama S Yamamoto K Sato S Prognostic factors of stage IV epithelial ovarian cancer: a multicenter retrospective study Gynecol Oncol 2001 81 398 403 11371128 10.1006/gyno.2001.6172
Davidson B Reich R Berner A Givant-Horwitz V Goldberg I Risberg B Kristensen GB Trope CG Bryne M Kopolovic J Nesland JM Ovarian carcinoma cells in serous effusions show altered MMP-2 and TIMP- 2 mRNA levels Eur J Cancer 2001 37 2040 2049 11597382 10.1016/S0959-8049(01)00235-0
Berner HS Davidson B Berner A Risberg B Kristensen GB Trope CG Van de Putte G Nesland JM Expression of CD44 in effusions of patients diagnosed with serous ovarian carcinoma – diagnostic and prognostic implications Clin Exp Metastasis 2000 18 197 202 11235996 10.1023/A:1006711320107
Davidson B Berner A Nesland JM Risberg B Berner HS Trope CG Kristensen GB Bryne M Ann Florenes V E-cadherin and alpha-, beta-, and gamma-catenin protein expression is up-regulated in ovarian carcinoma cells in serous effusions J Pathol 2000 192 460 469 11113863 10.1002/1096-9896(2000)9999:9999<::AID-PATH726>3.0.CO;2-M
Davidson B Berner A Nesland JM Risberg B Kristensen GB Trope CG Bryne M Carbohydrate antigen expression in primary tumors, metastatic lesions, and serous effusions from patients diagnosed with epithelial ovarian carcinoma: evidence of up-regulated Tn and Sialyl Tn antigen expression in effusions Hum Pathol 2000 31 1081 1087 11014575 10.1053/hupa.2000.9776
Davidson B Lazarovici P Ezersky A Nesland JM Berner A Risberg B Trope CG Kristensen GB Goscinski M van de Putte G Reich R Expression levels of the nerve growth factor receptors TrkA and p75 in effusions and solid tumors of serous ovarian carcinoma patients Clin Cancer Res 2001 7 3457 3464 11705863
Davidson B Goldberg I Givant-Horwitz V Nesland JM Berner A Bryne M Risberg B Kopolovic J Kristensen GB Trope CG van de Putte G Reich R Caveolin-1 expression in ovarian carcinoma is MDR1 independent Am J Clin Pathol 2002 117 225 234 11863219 10.1309/U40R-1BN4-6KJ3-BDG3
Davidson B Reich R Kopolovic J Berner A Nesland JM Kristensen GB Trope CG Bryne M Risberg B van de Putte G Goldberg I Interleukin-8 and vascular endothelial growth factor mRNA and protein levels are down-regulated in ovarian carcinoma cells in serous effusions Clin Exp Metastasis 2002 19 135 144 11964077 10.1023/A:1014582911680
Dokianakis DN Varras MN Papaefthimiou M Apostolopoulou J Simiakaki H Diakomanolis E Spandidos DA Ras gene activation in malignant cells of human ovarian carcinoma peritoneal fluids Clin Exp Metastasis 1999 17 293 297 10545015 10.1023/A:1006611220434
Chen Z Fadiel A Feng Y Ohtani K Rutherford T Naftolin F Ovarian epithelial carcinoma tyrosine phosphorylation, cell proliferation, and ezrin translocation are stimulated by interleukin 1alpha and epidermal growth factor Cancer 2001 92 3068 3075 11753986 10.1002/1097-0142(20011215)92:12<3068::AID-CNCR10149>3.0.CO;2-5
Scotton C Milliken D Wilson J Raju S Balkwill F Analysis of CC chemokine and chemokine receptor expression in solid ovarian tumours Br J Cancer 2001 85 891 897 11556842 10.1054/bjoc.2001.2020
Schwartz DR Kardia SL Shedden KA Kuick R Michailidis G Taylor JM Misek DE Wu R Zhai Y Darrah DM Reed H Ellenson LH Giordano TJ Fearon ER Hanash SM Cho KR Gene expression in ovarian cancer reflects both morphology and biological behavior, distinguishing clear cell from other poor- prognosis ovarian carcinomas Cancer Res 2002 2 6722 6726
Haviv I Campbell IG DNA microarrays for assessing ovarian cancer gene expression Mol Cell Endocrinol 2002 191 121 126 12044926 10.1016/S0303-7207(02)00063-1
Jazaeri AA Yee CJ Sotiriou C Brantley KR Boyd J Liu ET Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers J Natl Cancer Inst 2002 94 990 1000 12096084
Bayani J Brenton JD Macgregor PF Beheshti B Albert M Nallainathan D Karaskova J Rosen B Murphy J Laframboise S Zanke B Squire JA Parallel analysis of sporadic primary ovarian carcinomas by spectral karyotyping, comparative genomic hybridization, and expression microarrays Cancer Res 2002 62 3466 3476 12067990
Shridhar V Lee J Pandita A Iturria S Avula R Staub J Morrissey M Calhoun E Sen A Kalli K Keeney G Roche P Cliby W Lu K Schmandt R Mills GB Bast RC JrJames CD Couch FJ Hartmann LC Lillie J Smith DI Genetic Analysis of Early- versus Late-Stage Ovarian Tumors Cancer Res 2001 61 5895 5904 11479231
Schummer M Ng WV Bumgarner RE Nelson PS Schummer B Bednarski DW Hassell L Baldwin RL Karlan BY Hood L Comparative hybridization of an array of 21,500 ovarian cDNAs for the discovery of genes overexpressed in ovarian carcinomas Gene 1999 238 375 385 10570965 10.1016/S0378-1119(99)00342-X
Martoglio AM Tom BD Starkey M Corps AN Charnock-Jones DS Smith SK Changes in tumorigenesis- and angiogenesis-related gene transcript abundance profiles in ovarian cancer detected by tailored high density cDNA arrays Mol Med 2000 6 750 765 11071270
Hough CD Cho KR Zonderman AB Schwartz DR Morin PJ Coordinately up-regulated genes in ovarian cancer Cancer Res 2001 61 3869 3876 11358798
Schaner ME Ross DT Ciaravino G Sorlie T Troyanskaya O Diehn M Wang YC Duran GE Sikic TL Caldeira S Skomedal H Tu IP Hernandez-Boussard T Johnson SW O'Dwyer PJ Fero MJ Kristensen GB Borresen-Dale AL Hastie T Tibshirani R Van De Rijn M Teng NN Longacre TA Botstein D Brown PO Sikic BI Gene expression patterns in ovarian carcinomas Mol Biol Cell 2003 14 4376 4386 12960427 10.1091/mbc.E03-05-0279
Welsh JB Zarrinkar PP Sapinoso LM Kern SG Behling CA Monk BJ Lockhart DJ Burger RA Hampton GM Analysis of gene expression profiles in normal and neoplastic ovarian tissue samples identifies candidate molecular markers of epithelial ovarian cancer Proc Natl Acad Sci U S A 2001 98 1176 1181 11158614 10.1073/pnas.98.3.1176
Perou CM Sorlie T Eisen MB van de Rijn M Jeffrey SS Rees CA Pollack JR Ross DT Johnsen H Akslen LA Fluge O Pergamenschikov A Williams C Zhu SX Lonning PE Borresen-Dale AL Brown PO Botstein D Molecular portraits of human breast tumours Nature 2000 406 747 752 10963602 10.1038/35021093
Whitfield ML Sherlock G Saldanha AJ Murray JI Ball CA Alexander KE Matese JC Perou CM Hurt MM Brown PO Botstein D Identification of genes periodically expressed in the human cell cycle and their expression in tumors Mol Biol Cell 2002 13 1977 2000 12058064 10.1091/mbc.02-02-0030.
Lu Z Ghosh S Wang Z Hunter T Downregulation of caveolin-1 function by EGF leads to the loss of E-cadherin, increased transcriptional activity of beta-catenin, and enhanced tumor cell invasion Cancer Cell 2003 4 499 515 14706341 10.1016/S1535-6108(03)00304-0
Park W-Y Cho K-A Park J-S Kim D-I Park S Attenuation of EGF Signaling in Senescent Cells by Caveolin Annals of the New York Academy of Sciences 2001 928 79 84 11795531
Tibshirani R Hastie T Narasimhan Chu G Diagnosis of multiple cancer types by shrunken centroids of gene expression PNAS 2002 99 6567 6572 12011421 10.1073/pnas.082099299
Davidson B Goldberg I Gotlieb WH Kopolovic J Ben-Baruch G Nesland JM Berner A Bryne M Reich R High levels of MMP-2, MMP-9, MT1-MMP and TIMP-2 mRNA correlate with poor survival in ovarian carcinoma Clin Exp Metastasis 1999 17 799 808 11089877 10.1023/A:1006723011835
Goldberg I Davidson B Reich R Gotlieb WH Ben-Baruch G Berner A Nesland JM Kopolovic J aV integrin is a novel marker of poor prognosis in advanced-stage ovarian carcinoma Clin Cancer Res 2001 7 4073 4079 11751504
Davidson B Goldberg I Kopolovic J Gotlieb WH Givant-Horwitz V Nesland JM Berner A Ben-Baruch G Bryne M Reich R Expression of angiogenesis-related genes in ovarian carcinoma- A clinicopathologic study Clin Exp Metastasis 2000 18 501 507 11592307 10.1023/A:1011858225144
Luo LY Katsaros D Scorilas A Fracchioli S Bellino R van Gramberen M de Bruijn H Henrik A Stenman UH Massobrio M van der Zee AG Vergote I Diamandis EP The serum concentration of human kallikrein 10 represents a novel biomarker for ovarian cancer diagnosis and prognosis Cancer Res 2003 63 807 811 12591730
Luo LY Katsaros D Scorilas A Fracchioli S Piccinno R Rigault de la Longrais IA Howarth DJC Diamandis EP Prognostic value of Human Kallikrein 10 expression in epithelial ovarian carcinoma Clin Cancer Res 2001 7 2372 2379 11489815
Balleine RL Fejzo MS Sathasivam P Basset P Clarke CL Byrne JA The hD52 (TPD52) gene is a candidate target gene for events resulting in increased 8q21 copy number in human breast carcinoma Genes Chromosomes Cancer 2000 29 48 57 10918393 10.1002/1098-2264(2000)9999:9999<::AID-GCC1005>3.0.CO;2-O
Zhang S Lin Y Itaranta P Yagi A Vainio S Expression of Sprouty genes 1, 2 and 4 during mouse organogenesis Mech Dev 2001 109 367 70 11731251 10.1016/S0925-4773(01)00526-3
Yusoff P Lao DH Ong SH Wong ES Lim J Lo TL Leong HF Fong CW Guy GR Sprouty2 inhibits the Ras/MAP kinase pathway by inhibiting the activation of Raf J Biol Chem 2002 277 3195 201 11698404 10.1074/jbc.M108368200
Tamura S Hanada M Ohnishi M Katsura K Sasaki M Kobayashi T Regulation of stress-activated protein kinase signaling pathways by protein phosphatases Eur J Biochem 2002 269 1060 1066 11856337 10.1046/j.0014-2956.2002.02754.x
Camps M Nichols A Arkinstall S Dual specificity phosphatases: a gene family for control of MAP kinase function Faseb J 2000 14 6 16 10627275
O'Hagan RC Hassell JA The PEA3 Ets transcription factor is a downstream target of the HER2/Neu receptor tyrosine kinase Oncogene 1998 16 301 310 9467955 10.1038/sj.onc.1201547
Davidson B Goldberg I Gotlieb WH Kopolovic J Ben-Baruch G Reich R PEA3 is the second Ets family transcription factor involved in tumor progression in ovarian carcinoma Clin Cancer Res 2003 9 1412 1419 12684413
Givant-Horwitz V Davidson B Lazarovici P Schaefer E Nesland JM Trope' CG Reich R Mitogen-activated protein kinases (MAPK) as predictors of clinical outcome in serous ovarian carcinoma in effusions Gynecol Oncol 2003 91 160 172 14529677 10.1016/S0090-8258(03)00434-7
Givant-Horwitz V Davidson B Goderstad JM Nesland JM Trope' CG Reich R The PAC-1 dual specificity phosphatase predicts poor outcome in serous ovarian carcinoma Gynecol Oncol 2004 93 517 523 15099972 10.1016/j.ygyno.2004.03.009
Davidson B Malignant effusions: from diagnosis to biology. Diagn Cytopathol 2004 31 246 254 15452897
Davidson B Givant-Horwitz V Lazarovici P Risberg B Nesland JM Trope CG Schaefer E Reich R Matrix metalloproteinases (MMP), EMMPRIN (extracellular matrix metalloproteinase inducer) and mitogen-activated protein kinases (MAPK): co-expression in metastatic serous ovarian carcinoma Clin Exp Metastasis 2003 20 621 31 14669793 10.1023/A:1027347932543
Chang HY Sneddon JB Alizadeh AA Sood R West RB Montgomery K Chi JT van de Rijn M Botstein D Brown PO Gene Expression Signature of Fibroblast Serum Response Predicts Human Cancer Progression: Similarities between Tumors and Wounds PLoS Biol 2004 2 E7 14737219 10.1371/journal.pbio.0020007
Tang TC Sham JS Xie D Fang Y Huo KK Wu QL Guan XY Identification of a candidate oncogene SEI-1 within a minimal amplified region at 19q13.1 in ovarian cancer cell lines Cancer Res 2002 62 7157 61 12499249
Thompson FH Nelson MA Trent JM Guan XY Liu Y Yang JM Emerson J Adair L Wymer J Balfour C Massey K Weinstein R Alberts DS Taetle R Amplification of 19q13.1-q13.2 sequences in ovarian cancer. G-band, FISH, and molecular studies Cancer Genet Cytogenet 1996 87 55 62 8646743 10.1016/0165-4608(95)00248-0
Davidson B Risberg B Kristensen G Kvalheim G Emilsen E Bjamer A Berner A Detection of cancer cells in effusions from patients diagnosed with gynaecological malignancies. Evaluation of five epithelial markers Virchows Arch 1999 435 43 49 10431845 10.1007/s004280050393
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 JM The Stanford Microarray Database Nucleic Acids Res 2001 29 152 155 11125075 10.1093/nar/29.1.152
Tusher VG Tibshirani R Chu G Significance analysis of microarrays applied to the ionizing radiation response Proc Natl Acad Sci U S A 2001 98 5116 5221 11309499 10.1073/pnas.091062498
Troyanskaya O Cantor M Sherlock G Brown P Hastie T Tibshirani R Botstein D Altman RB Missing value estimation methods for DNA microarrays Bioinformatics 2001 17 520 525 11395428 10.1093/bioinformatics/17.6.520
Greene GF Kitadai Y CPettaway A von Eschenbach AC Bucana CD Fidler IJ Correlation of metastasis-related gene expression with metastatic potential in human prostate carcinoma cells implanted in nude mice using an in situ messenger RNA hybridization technique Am J Pathol 1997 150 1571 82 9137084
Xing X Wang SC Xia W Zou Y Shao R Kwong KY Yu Z Zhang S Miller S Huang L Hung MC The ets protein PEA3 suppresses HER-2/neu overexpression and inhibits tumorigenesis Nat Med 2000 6 189 195 10655108 10.1038/72294
Davidson B Goldberg I Gotlieb WH Kopolovic J Ben-Baruch G Nesland JM Berner A Bryne M Reich R High levels of MMP-2, MMP-9, MT1-MMP and TIMP-2 mRNA correlate with poor survival in ovarian carcinoma Clin Exp Metastasis 1999 17 799 808 11089877 10.1023/A:1006723011835
Reed JA Manahan LJ Park CS Brigati DJ Complete one-hour immunocytochemistry based on capillary action Biotechniques 1992 13 434 43 1389176
|
16042759
|
PMC1236614
|
CC BY
|
2021-01-04 16:36:35
|
no
|
Mol Cancer. 2005 Jul 21; 4:26
|
utf-8
|
Mol Cancer
| 2,005 |
10.1186/1476-4598-4-26
|
oa_comm
|
==== Front
Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-611609553510.1186/1743-422X-2-61ResearchDetection, quantification and genotyping of Herpes Simplex Virus in cervicovaginal secretions by real-time PCR: a cross sectional survey Aryee Esther AN [email protected] Robin L [email protected] Angels [email protected] Steve [email protected] Martin J [email protected] Medical Research Council Laboratories, Fajara, The Gambia2 London School of Hygiene and Tropical Medicine, London, UK2005 11 8 2005 2 61 61 17 6 2005 11 8 2005 Copyright © 2005 Aryee et al; licensee BioMed Central Ltd.2005Aryee 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
Herpes Simplex Virus (HSV) Genital Ulcer Disease (GUD) is an important public health problem, whose interaction with HIV results in mutually enhancing epidemics. Conventional methods for detecting HSV tend to be slow and insensitive. We designed a rapid PCR-based assay to quantify and type HSV in cervicovaginal lavage (CVL) fluid of subjects attending a Genito-Urinary Medicine (GUM) clinic. Vaginal swabs, CVL fluid and venous blood were collected. Quantitative detection of HSV was conducted using real time PCR with HSV specific primers and SYBR Green I. Fluorogenic TaqMan Minor Groove Binder (MGB) probes designed around a single base mismatch in the HSV DNA polymerase I gene were used to type HSV in a separate reaction. The Kalon test was used to detect anti-HSV-2 IgG antibodies in serum. Testing for HIV, other Sexually Transmitted Infections (STI) and related infections was based on standard clinical and laboratory methods.
Results
Seventy consecutive GUM clinic attendees were studied. Twenty-seven subjects (39%) had detectable HSV DNA in CVL fluid; HSV-2 alone was detected in 19 (70%) subjects, HSV-1 alone was detected in 4 (15%) subjects and both HSV types were detected in 4 (15%) subjects. Eleven out of 27 subjects (41%) with anti-HSV-2 IgG had detectable HSV-2 DNA in CVL fluid. Seven subjects (10%) were HIV-positive. Three of seven (43%) HIV-infected subjects and two of five subjects with GUD (40%) were secreting HSV-2. None of the subjects in whom HSV-1 was detected had GUD.
Conclusion
Quantitative real-time PCR and Taqman MGB probes specific for HSV-1 or -2 were used to develop an assay for quantification and typing of HSV. The majority of subjects in which HSV was detected had low levels of CVL fluid HSV, with no detectable HSV-2 antibodies and were asymptomatic.
==== Body
Background
Genital herpes, which is caused mainly by Herpes Simplex Virus (HSV) -2 [1] but also by HSV-1 [2] remains a worldwide problem [3]. The strongest known risk factor for the heterosexual transmission of Human Immunodeficiency Virus (HIV) and other Sexually Transmitted Infections (STI) is Genital Ulcer Disease (GUD) [4]. Over the past decade, HSV-2 has been identified as the most common aetiological agent of GUD [5]. Studies of HSV-2 seroprevalence have found high rates in African-Americans [6] and in African populations in Uganda, Zimbabwe, Tanzania, Central African Republic, South Africa and The Gambia [7-12]. In The Gambia, HSV-2 seropositivity among young adults from rural communities was 28% in women and 5% in men which increased with age [12]. HSV-2 seroprevalence increases with high risk sexual behaviour [9] and with factors related to polygynous marriage practices in rural populations [13]. The majority of subjects infected with HSV-2 are asymptomatic but exhibit sub-clinical cervicovaginal virus secretion which is thought to be important in the transmission of HSV-2 [14,15].
Antiviral therapy with acyclovir or valacyclovir, used during episodes of primary and recurrent HSV-2 GUD, reduces both the rate of secretion, and the rate at which GUD develops [16,17]. These drugs were also found to reduce the transmission from an infected person to another susceptible individual [18] and to minimise sub-clinical HSV-2 genital secretion, preventing the spread of disease [19]. Resistance to antiviral drugs has been reported but occurs infrequently [20]. Interventions therefore can be helpful in the reduction of disease by preventing the spread of HSV, which consequently impacts on HIV transmission. The effectiveness of such interventions demands rapid, efficient, reliable and type specific assays. These assays can serve as biological endpoints in deciding when to administer intervention, monitoring the effectiveness of any current intervention, determining the efficacy of drugs, assessing drug resistance and are useful research tools in the study of the epidemiology of transmission.
The discrimination of HSV-1 from HSV-2 was originally performed using virus culture followed by antibody binding to type-specific determinants (virus neutralisation) [21]. The application of molecular methods, such as restriction fragment length polymorphism (RFLP) analysis of HSV PCR amplicons is thought to provide a reliable method of typing the virus [21]. Whilst serology based typing methods target surface exposed epitopes such as those on glycoprotein C or G, molecular typing has largely exploited differences between HSV-1 and -2 DNA polymerase I genes. Archetypal HSV-1 and -2 DNA polymerase I genes share 93% sequence identity and 82% amino acid homology. The selection of strain typing polymorphisms for molecular methods is based on the sequence information deposited in public databases coupled with the availability of a convenient restriction endonuclease site. This can identify variation at a selected single nucleotide polymorphism (SNP) site. A rapid SNP typing method is useful because it can yield information about the virus population in the affected host population. This is of value in classification and in epidemiological studies aimed to investigate host-pathogen interplay.
A more efficient method for diagnosis of HSV infection is to use PCR in real time for detection and quantification. HSV SNP sub-typing by 'allele' specific fluorogenic probes offers many advantages over RFLP methods or viral culture. Amplification of the target DNA, and hybridization to a fluorogenic probe are conducted in a single PCR and therefore the chances of possible contamination are minimised. The main advantage of real-time detection is the large dynamic range offered in a quantitative assay coupled to the ability to discriminate between fluorophores in a multiplex reaction. We selected Taqman probes incorporating the minor groove binder (MGB), 1,2-dihydro-(3H)-pyrrolo [3,2-e] indole-7-carboxylate (CDPI3) [22]. MGB probes offer high sensitivity and accuracy, due to their short length which increases the sensitivity and stability of probe-sequence complexes to single base changes [23]. However, important consideration should be given to the selection of the SNP under investigation. Recent work using Eclipse-MGB probes which bind to a highly polymorphic region of HSV glycoprotein D found that sequence polymorphisms in the probe binding region decreased the sensitivity of typing assay [24]. The present study used assays based on Taqman-MGB probes, to identify the HSV type in a population of symptomatic and asymptomatic patients attending a GUM clinic in The Gambia. The possible role of other co-infections in the secretion of HSV in CVL fluid and HSV transmission were investigated.
Results
Study subjects
Seventy subjects included in the study were of median age 27 years (range 17–50). Genital examination revealed that 5/70 subjects had GUD (four external and one cervical). Four subjects withheld consent for HIV serology. There was one known HIV positive patient identified at a previous Out Patient Department visit. Two further samples were not tested for HSV-2 IgG (total tested n = 63) and one sample was not tested for Hepatitis B and Treponema pallidium (n = 62) because of insufficient sample volume.
Quantitative analysis of HSV viral load in CVL
Amplification of the HSV DNA and hybridization to a fluorogenic probe were conducted in different PCR reactions. Three μl of extracted template DNA of a 200 μl eluate prepared from CVL was used. Most samples had values less than 10 viral copies/PCR reaction but were positive as indicated by the presence of an amplification peak of the correct Tm in dissociation or melting curve analysis plots. The accurate estimation of the quantity of HSV was expressed as viral copies/ml of CVL. The lower limit of quantitation of the HSV assay was therefore 335 viral copies/ml cervical lavage fluid. Forty-three samples had no detectable HSV amplicons. Twenty-seven out of 70 subjects (39%) had HSV detected in CVL fluid. Figure 1. shows the HSV load in the CVL fluid of these 27 HSV PCR positive subjects. The distribution is left censored and the majority of subjects had <335 copies/ml of lavage. HSV secretion by quantitative real time PCR in positive subjects ranged from < 335 to 10,409,000 viral genome copies/ml of CVL fluid.
Determination of HSV-1 and -2 types by Taqman-MGB probes
Nineteen of the 27 HSV positive subjects (70%) were identified as HSV-2, 4/27 (15%) were HSV-1 and 4/27 (15%) were positive for both HSV-1 and -2 by HSV probe specific binding assay. HSV type was confirmed in 7 samples in which an equivocal typing result was initially obtained. HSV type was confirmed by repeating the real-time assay using the Rotorgene 3000 instrument (Corbett Research Ltd, Sydney, Australia). Sequencing of all PCR amplicons and reference control strains gave a 100% confirmation with that of the probe at the SNP site (Figure 1).
HSV type and serology
Anti HSV-2 IgG antibodies were detected in 29/63 (46%) subjects, most of whom were not secreting HSV. The subject with the highest number of HSV-2 viral genome copies/ml of CVL fluid had no detectable anti HSV-2 IgG. Two other CVL fluid samples in which high levels of HSV DNA were detected were typed as HSV-2 and these were positive for anti HSV-2 IgG antibodies. In total 10/21 subjects in whom HSV-2 secretion was detected were negative for HSV-2 antibodies, whilst 11/21 subjects in which HSV-2 secretion was detected were HSV-2 antibody positive. Two subjects in which HSV-2 secretion was detected were unable to be tested for anti HSV-2 IgG due to insufficient serum. A single HSV-1 secretion positive subject was positive for anti HSV-2 IgG and 3 HSV-1 secretion positive samples were negative for anti HSV-2 IgG antibodies.
Potential risk factors which may be associated with HSV cervicovaginal secretion
Possible cofactors for genital HSV infection were examined (Table 1) but none of these were associated with current HSV secretion in CVL fluid. Data in Table 2 further demonstrates the lack of a relationship between each risk co-factor and the quantity of HSV present in CVL. Eighteen (67%) subjects judged to be secreting HSV had < 335 viral copies / ml of CVL fluid. Five of 70 subjects had GUD, 2 of 5 were currently secreting HSV (HSV-2 by typing and positive for anti-HSV-2 IgG). Seven of 66 subjects (11%) were found to be HIV-1 seropositive. Thirty-five out of 70 (50%) were diagnosed as having Bacterial Vaginosis (BV). Twenty-nine of 70 (41%) subjects had Candida and Trichomonas vaginalis was found in 7 out of 70 (10%). Eight out of 62 (13%) were positive for Hepatitis B and 7 out of 70 (10%) were positive for Chlamydia trachomatis pgp3. None of the subjects were diagnosed as having Neisseria gonorrhoea, or clue cells. Two of 62 (3%) subjects were diagnosed as T. pallidium infected, neither of which secreted HSV.
Table 1 Association between potential risk factors and HSV detected in CVL
Cofactors Present Absent Relative risk (95% CI) P-value
GUD 2/5 (40%) 25/65 (38%) 1.04 (0.34 – 3.18) 0.68
Anti HSV-2 IgG* (Kalon test) 11/27 (41%) 7/32 (22%) 1.86 (0.84 – 4.13) 0.19
HIV 3/7 (43%) 23/59 (39%) 1.1 (0.44 – 2.74) 0.83
BV 14/35 (40%) 13/35 (37%) 1.08 (0.60 – 1.95) 1.00
C. trachomatis 5/7 (71%) 22/63 (35%) 2.05 (1.15 – 3.64) 0.14
Hepatitis B 5/8 (63%) 22/54 (41%) 1.53 (0.82 – 2.87) 0.44
Candida 9/29 (31%) 18/41 (44%) 0.71 (0.37 – 1.35) 0.40
T. vaginalis 3/7 (43%) 24/63 (38%) 1.13 (0.45 – 2.80) 0.86
* Comparison with HSV-2 only. P values calculated by χ2 test
Table 2 Relationship of HSV CVL viral load with potential risk factors
Cofactor Geometric mean number of copies of virus/ml of lavage fluid
Present Absent P-values
GUD 8900 <335 0.1532
Anti HSV-2 IgG 1100* <335* 0.2976
HIV-1 infection <335 <335 0.5568
BV 400 <335 0.5020
Candida <335 <335 0.8320
Hepatitis B 400 <335 0.2480
C. trachomatis <335 <335 0.1890
T. vaginalis 17100 <335 0.0760
*Comparison done with HSV-2 only. P values calculated by non-parametric Kruskal-Wallis test.
Determination of HSV-1 and -2 types by Tm
The melting temperatures (Tm) of the HSV amplicon obtained from the dissociation or melting curve plot following amplification and quantification of HSV are shown in Figure 1. Both positive and control HSV-1 samples had Tm ranging from 86.6°C to 86.9°C. HSV-2 positive samples, ranged from 87.0 – 88.0 °C and an example of the dissociation plot is shown in Figure 2. Dual HSV -1 and -2 infections, confirmed by sequencing, also had a Tm range between 87.0–88.0°C. Of note in the sequences is the high number of mutations or alternative bases contained with in the probe binding area. Surprisingly probe binding and Tm appear largely unaffected by these changes (Figure 1). Of the 7 samples confirmed by Rotorgene, 4 of these had no mutations in the probe binding area. The remaining 3 samples resulted in poor quality sequencing reactions and the results were not interpretable.
Figure 1 R = A/G; M = A/C; N = A/T/G/C. Control HSV-1 and -2 DNA was obtained from 2003 Quality Control Molecular Diagnostics 2003 Proficiency panel (Block 6, Kelvin Campus, West of Scotland Science Park, Glasgow UK). The reference sequences were based on Blast results from NCBI. The Tm of the PCR product (146 b.p.) which was amplified during quantitation with SYBR Green I is shown against its complementary sequence with the SNP position marked in bold. The probe and sequence ascertainment of types were in agreement. * Tm was unable to distinguish dual infection in this assay. -n.c. = not confirmed by sequencing.
Figure 2 Dissociation curves of amplicons used to identify HSV-1, HSV-2 and dual positive samples. No curves were observed in HSV negative subjects. Tm was recorded and compared with probe binding and sequence results. Collectively nine different peaks with Tm in the range 86.6 – 88.0°C could be observed. Three Tm representative of HSV-1, HSV-2 and dual positive samples confirmed by sequencing are shown with Tm indicated.
Discussion
Diagnostic methods, ranging from traditional culture and serological detection methods, to molecular techniques, have been described for the diagnosis of HSV. Most of these assays, including the gold standard of viral isolation by culture [25] are slow and prone to contamination. The assay turn-around time for culture is 4 days as compared to that of 4 hours for enzyme immunoassay (EIA) and 2–4 hours for real-time PCR [26]. Viral culture diagnosis is useful if HSV-2 is responsible for symptomatic infection in the form of vesicles or ulcers, when live virus can usually be isolated. Success of detection further depends on the secretion of virus during sampling. Its sensitivity relies on the way samples are collected, transported and stored [21]. Cell culture can only be done in laboratories with expertise and facilities; in developing countries this facility may not be available. Accurate serological tests are appropriate in asymptomatic cases, when viral culture and PCR assays are largely negative [15].
Several commercially available HSV-type specific serological assays are available, but a test such as a HSV-2 Western blot is expensive and restricted largely to reference laboratories. The Kalon test, which was found to be the best among a set of serological tests evaluated in samples from different African cities (with sensitivity and specificity of 92.3% and 97.7% respectively) [27], tests only latent infection and may not detect recent seroconversion [28]. DNA amplification using PCR techniques is reported to be more sensitive than culture, and a number of studies have used fluorescent based real time PCR techniques with primers targeting sequences from HSV glycoprotein B, thymidine kinase or DNA polymerase genes [21,29]. Some PCR assays require laborious post-PCR procedures such as RFLP analysis [21], which may introduce a risk of contamination. The high degree of sequence homology between HSV-1 and -2 makes the design of type specific primers challenging [30], nevertheless this has been attempted with varying success, along with Amplification Refractory Mutation System PCR [31].
Our assays were able to estimate HSV load and distinguished specific HSV-1 and -2 cervicovaginal viral secretion. Amplification and typing could not be carried out in a single PCR because, under the conditions used, insufficient specific amplicons were generated for accurate typing in a single step. This may have circumvented problems relating to sensitivity of the HSV-1 and -2 probe relative binding. In a single step multiplex assay mutations in the probe binding area are reported to lead to a loss in sensitivity and error in the classification of samples [24]. When diluted amplified HSV amplicons were used for a further typing reaction, HSV-2 was more commonly detected than HSV-1 (70% opposed to 15%). This is concordant with recent work that found most subjects were secreting HSV-2 [32]. An earlier study of Gambian commercial sex workers found that 26% of women were secreting HSV but the study could not distinguish HSV strain types (Aryee et al unpublished observation). In the current study most of the subjects secreting HSV were of age ranging 20 – 41 years. This confirms earlier studies in which HSV-2 was most prevalent (15 – 34 year old subjects from rural Gambian communities) [12]. Thus Gambian women in their twenties appear at highest risk of HSV-2 infection. Most of the women reported in this study were found to be secreting low levels of viral DNA in CVL fluid. Anti HSV-2 IgG was detected in 29 out of 63 (46%) subjects which is higher than a previous Gambian study by Shaw et al [12]. The subjects for that study were from rural Gambian communities whereas our work was with GUM clinic attendees, whose risk of HSV-2 infection is greater than the general population [33]. HIV has been found to enhance the expression of HSV-2 [11]. However, it is unlikely that co-infection with HIV is responsible for the increased HSV-2 sero-prevalence in this study given the low rate of HIV infection among the study subjects and the low level of HSV secretion in HIV-1 positive subjects compared to HIV-1 negative subjects.
We found anti HSV-2 IgG seropositivity correlated poorly with HSV-2 secretion and several factors may have contributed to this observation. IgG seropositivity may take time to develop and the Kalon antibody test may not be sensitive enough to detect early seroconversion. HSV-2 could therefore be present in secretions without established sero-conversion. It is known that HSV-2 genital secretion is intermittent even in HSV seropositive subjects so it is possible that these subjects have only recently been exposed such that a detectable immune response has not yet developed. The highest HSV-2 viral load in lavage fluid was found in a seronegative subject. This result may not be conflicting if this was a newly acquired infection and only a primary antibody (IgM) response was stimulated with levels of IgG below the sensitivity of the Kalon test. Follow-up of subjects is required to investigate whether subjects in which HSV secretion was identified but were seronegative for anti HSV-2 IgG have now seroconverted.
The data suggest that while STI such as BV, HIV, C. trachomatis and Hepatitis B may increase the rate and quantity of HSV-2 secretion, these effects were not statistically significant which is probably due to the low numbers of women recruited for the study. The study suggested that subjects with GUD tend to secrete more virus than those without GUD, however, most subjects that were secreting virus were asymptomatic and did not have GUD in line with earlier work [16,17]. The melting temperature (Tm) of PCR amplicons has been used to identify HSV types by others [24,32] but these methods could not distinguish between dual and mono specific HSV-2 infection. The use of amplicon Tm for the assignment of a genotype has been utilized for human SNPs using High-Resolution melting instruments [34]. Whilst this may be applicable for the relatively stable sequences in the human genome it may not be a sustainable method for highly changeable viral sequences as suggested by others.
The T to C (A/G in the reverse strand) transition that we have identified as a distinguishing SNP appears to be indicative of either HSV-1 or HSV-2, however, there are a limited number of HSV sequences available in public sequence databases to indicate that every HSV-1 or 2 will have either T or C at that position. It has not been demonstrated that these SNPs, PCR-RFLPs or Tm correlate with monoclonal type specific antibody reactivity or unique region sequence data. Further data need to be gathered to evaluate the usefulness of these methods and their application to population and epidemiological studies.
Conclusion
This assay was able to distinguish HSV-1 from HSV-2 and quantify HSV genital secretion. Thirty nine percent of women attending the GUM clinic were secreting HSV and most of these had low viral loads in CVL fluid with no detectable anti-HSV-2 IgG antibodies and were asymptomatic. The presence of other STI may facilitate HSV secretion but further studies with a larger sample size are required to investigate whether the HSV type or whether low levels of HSV genital secretion are important in the transmission of infection.
Methods
Subjects
Seventy consecutive female subjects, attending the GUM clinic at MRC Fajara, The Gambia from April to June 2004 were recruited. After giving informed consent, clinical data about the subjects were recorded. This was conducted by questionnaire and an examination for genital lesion by the clinician/nursing officer. The study was approved and conducted under the guidelines of The Gambian Government and MRC Joint Ethics Committee.
Specimens
Two vaginal/cervical swab specimens were taken, after which the cervicovaginal area was flushed with 10 ml of phosphate buffered saline (PBS) for 1 minute and aspirated into sterile tubes. Samples were kept on ice and transported promptly to the laboratory. One ml of venous blood was also collected and allowed to clot before centrifugation at 800 × g for 10 minutes to isolate serum. Serum was stored at -20°C until used.
Processing specimens
Cervicovaginal lavage samples were centrifuged at 1000 × g for 10 min and the supernatant discarded. Cellular materials were resuspended in 1 ml PBS and stored at -70°C. A high vaginal swab was used to make a smear on clean slides for Gram staining. The second swab was used for routine microbiological analysis of STI.
DNA extraction
DNA was extracted from 200 μl of lavage cell suspension using the QiaAmp DNA Mini kit (QIAGEN Ltd, Crawley, UK) according to manufacturers instructions.
Selection of HSV typing single nucleotide polymorphism
A survey of HSV-1 and -2 DNA polymerase I gene sequences available through the National Center for Biotechnology Information (NCBI) was conducted. Eight HSV-1 [EMBL:X03181.1], [EMBL:X04495.1], [EMBL:X04771.1], [EMBL:X14112.1], [DDBJ:AB072389.1], [DDBJ:AB070848.2], [DDBJ:AB070847.2], [GenBank:M10792.1] and 5 HSV-2 [GenBank:AY038367.1], [EMBL:Z86099.2], [GenBank:M14793.1], [GenBank:M16321.1], [GenBank:AY038366.1] sequences were identified. Following alignment, candidate SNPs were selected in regions with no other base changes within 20 nucleotides (i.e. within the likely probe binding area) of the potential typing SNP. These SNPs were then submitted for primer-probe design using either Primer Express v2.0 (Applied Biosystems Inc, Warrington, UK) or using the web based service of Epoch Biosciences . The optimum primer-probe design combination was then selected for synthesis.
Detection of HSV DNA, probe typing & melting point determination
Quantitative PCR was performed on the ABI 5700 sequence detection system (Applied Biosystems Inc, Warrington, UK). Duplicate 3 μl samples of extracted DNA were added to a 22 μl PCR master mix (PCR SYBR Green I, QIAGEN Ltd, Crawley, UK) containing 0.4 μM each primer. The primers amplified a generic HSV 146 b.p. product from the HSV DNA polymerase I gene. [forward primer – 5'-AGCCTGTACCCCAGCATCAT-3'; reverse primer – 5'-TGGGCCTTCACGAAGAACA-3']. Cycling temperatures were 95°C for 15 minutes, followed by 40 cycles of 94°C for 15s, 58°C for 30s and 72°C for 30s. At the end of amplification PCR products were subjected to a dissociation or melting curve analysis and the Tm of the peak was recorded. HSV PCR products were diluted one in ten with DNase free RNase free water and probe typed using Taqman MGB probes (Applied Biosystems, Inc, Warrington, UK) directed against the HSV DNA polymerase I gene. These probes specifically detected a C/T SNP at position 2202 (C) for HSV-1, and 2451 (T) for HSV-2. The probe sequences were: HSV-1 (5'-VIC-AGCGTgCTGAAGC-MGB-Q-3') and HSV-2 (5'-6FAM-AGAGCGTaCTGAAGCA-MGB-Q-3'). The probes were used in a single tube real time PCR using the QuantiTect probe kit (QIAGEN Ltd, Crawley, UK) with the following cycling conditions using either an Opticon 2 (GRI/MJ Research, Braintree, UK) or Rotorgene 3000 (Corbett Research, Sydney, Australia) thermal cycler: 95°C for 15 minutes, followed by 40 cycles of 94°C for 15s, 68°C for 30s and 76°C for 30s. Fluorescence was acquired at the end of the annealing phase. For the Rotorgene 3000 HSV typed samples, an annealing phase of 66°C for 30s was used. The quantitative assays performed on the ABI sequence detection system included standards of 106 to 10 copies per reaction and negative controls. These were used to generate a standard curve and calculate the copy number of the unknown samples. HSV positive samples from the Quality Control Molecular Diagnostics 2003 Proficiency Panel (Block 6, Kelvin Campus, West of Scotland Science Park, Glasgow UK) were used as positive controls. All DNA samples were tested for inhibition of PCR using bacteriophage lambda (λ) DNA and primers. Briefly, test samples were 'spiked' into a PCR reaction containing approximately 100 copies of bacteriophage λ DNA and a primer pair directed against λ DNA. The performance of the PCR was monitored by quantitative real-time PCR (qPCR). The mean cycle threshold (Ct) and the standard deviation of the controls were calculated. Samples in which the mean Ct of the test sample fell outside the mean Ct plus three standard deviations of the controls, were judged to be inhibitory. Inhibitory samples were re-extracted by a repeat of the QiaAmp Mini kit extraction method and retested in the qPCR and inhibition assays.
HSV amplicon sequence confirmation of probe typing
One in ten dilutions of the amplified positive products were prepared using DNase free RNase free water. A PCR reaction was prepared by adding 6 μl of the diluted amplified positive products to 44 μl of a PCR Hotstar Taq master mix (QIAGEN Ltd, Crawley, UK) and the HSV primers with the addition of M13 primer sequences [HSV-M13 forward 5'-TGTAAAACGACGGCCAGTAGCCTGTACCCCAGCAT-3'; HSV-M13 reverse 5'-CAGGAAACAGCTATGACCTGGGCCTTCACGAAGA-3']. Cycling temperatures were the same as for the HSV real-time quantitative assay, which used SYBR Green I, modified by the addition of 5 extra cycles and a final extension at 72°C for 5 min. PCR DNA product and purity were checked by electrophoresis using a 2% agarose gel. PCR products with no primer-dimers present were purified using Qiagen DNA mini-kits (QIAGEN Ltd, Crawley, UK). When primer-dimers were observed specific PCR amplicons were gel purified (QIAGEN Ltd, Crawley, UK). Purified PCR products were then sent to the Wellcome Trust Centre for Human Genetics, Oxford, UK for dye-primer Sanger sequencing on an ABI 3100 (Applied Biosystems, Inc, Warrington, UK) capillary automated sequencer.
Detection of Chlamydia trachomatis pgp3 gene
The presence of C. trachomatis DNA was detected and quantified by Quantitect SYBR Green I on the ABI 5700 sequence detection system using C. trachomatis pgp3 primers [forward primer 5'-GATGCGGAAAAAGCTTACCA-3'; reverse primer 5'-TGAATAACCCGTTGCATTGA-3']. These primers amplified a product of 193 b.p. from the multicopy cryptic chlamydial plasmid. PCR cycling conditions were as recommended by the manufacturer annealing at 59°C for 30s and extension at 72°C for 20s for 40 cycles. Standards of 106 to 10 copies per reaction of C. trachomatis pgp3 amplicons and negative controls were included in each PCR reaction to generate a standard curve and quantities of the unknown samples estimated as before.
Serology
Serum anti HSV-2 IgG was detected using the Kalon IgG kit (Kalon Biologicals, Ashgate, UK) and followed the manufacturer's instructions. Detection of antibodies to HIV-1 and HIV-2 in serum was done using Murex ICE HIV 1.2.0 ELISA Test kit (Murex, Dartford, Kent, UK). Reactive samples were then subjected to further testing using Monospecific ELISA, Murex ICE HIV-2 for HIV-2 diagnosis and Wellcozyme HIV Recombinant for HIV-1 (Murex, Dartford, Kent, UK). Diagnoses were confirmed on a second serum sample collected two weeks after the first sample. For Hepatitis B the Abbott Determine™ (Abbott Laboratories, Illinois, USA) HBsAg qualitative immunoassay was used to detect Hepatitis B surface Antigen (HBsAg) in serum samples by following the manufacturer's instructions. Serum samples from patients were also screened for T. pallidum using MACRO-VUE Rapid plasma Reagin (BD Biosciences, Oxford, UK) test kit and following manufacturer's protocol. Positive samples were confirmed using a T. pallidum haemagglutination assay, Micro syph TP-200 (Axis-Shield Diagnostics LTD, Huntingdon, UK).
Microbiology
Gram stained slides were observed for the presence or absence of Lactobacilli, BV associated organisms, Mobiluncus and clue cells. Diagnosis of BV was based on the Nugent Score. Cervical swabs were used to make smears on slides, Gram stained and observed for Gram-negative intracellular diplococci. Culture for the isolation of N. gonorrhoea was performed on Thayer Martin's medium supplemented by vitox. Any positive cultures were tested for oxidase and carbohydrate oxidation as confirmation of N. gonorrhoea.
Candida, Trichomonas vaginalis and clue cells
A few drops of saline was used to make a wet preparation of the high vaginal swab and observed under a light microscope for the presence of Candida, T. vaginalis and clue cells (granulated epithelial cells with Gardnerella vaginalis attached).
Statistical analysis
HSV viral copy numbers were log transformed before statistical analysis. Statistical analysis was carried out in EPI Info, SPSS and Minitab. Kruskal-Wallis and χ2 tests were used as indicated in the results.
List of Abbreviations
CDPI3: tripeptide 1,2-dihydro-(3H)-pyrrolo [3,2-e]indole-7-carboxylate
Ct: Cycle threshold
CVL: Cervicovaginal lavage
DNA: Deoxyribonucleic acid
EIA: Enzyme immunoassay
ELISA: Enzyme-Linked Immunosorbent Assay.
gG: glycoprotein G
GUD: Genital Ulcer Disease
GUM: Genito-Urinary Medicine
HBsAg: Hepatitis B surface Antigen
HIV: Human Immunodeficiency Virus
HSV: Herpes Simplex Virus
Ig: Immunoglobulin
MAb: Monoclonal antibodies
MGB: Minor Groove Binder
MRC: Medical Research Council
PBS: Phosphate Buffered Saline
PCR: Polymerase Chain Reaction
qPCR: quantitative Polymerase Chain Reaction
SNP: Single Nucleotide Polymorphism
STD: Sexually Transmitted Disease
STI: Sexually Transmitted Infection
Tm Melting temperature
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
The study was designed by MJH and EANA; experimental work was done by EANA, MJH and AN; interpretation and laboratory work was conducted by MJH, EANA, AN, SK and RB; EANA, RLB, SK and MJH were responsible for analysis of results and preparation of the manuscript.
Acknowledgements
The authors wish to thank Dr Beryl West, CDC Uganda and Dr Sam McConkey, MRC Laboratories, Gambia for routine reagents, helpful advice and discussions. We also thank clinical staff of the GUM clinic, serology and microbiology/reproductive health at MRC, The Gambia. We thank Ms Sarah Burl for critical review of the manuscript. The study was supported by funds from the MRC training committee and grants from the MRC UK. Finally we thank the study participants.
==== Refs
O'Farrell N Increasing prevalence of genital herpes in developing countries: implications for heterosexual HIV transmission and STI control programmes Sex Transm Infect 1999 75 377 384 10754939
Kinghorn GR Herpes simplex Type 1 Genital Infections Herpes 1999 6 4 7
Nahmias AJ Lee FK Beckman-Nahmias S Sero-epidemiological and -sociological patterns of herpes simplex virus infection in the world Scand J Infect Dis Suppl 1990 69 19 36 2175939
Greenblatt RM Lukehart SA Plummer FA Quinn TC Critchlow CW Ashley RL D'Costa LJ Ndinya-Achola JO Corey L Ronald AR Genital ulceration as a risk factor for human immunodeficiency virus infection Aids 1988 2 47 50 3128996
Austin H Macaluso M Nahmias A Lee FK Kelaghan J Fleenor M Hook EW Correlates of herpes simplex virus seroprevalence among women attending a sexually transmitted disease clinic Sex Transm Dis 1999 26 329 334 10417020
Wald A Herpes Simplex Virus Type 2 Transmission: Risk Factors and Virus Shedding Herpes 2004 11 130A 137A 15319082
Wagner HU Van Dyck E Roggen E Nunn AJ Kamali A Schmid DS Dobbins JG Mulder DW Seroprevalence and incidence of sexually transmitted diseases in a rural Ugandan population Int J STD AIDS 1994 5 332 337 7819350
Gwanzura L McFarland W Alexander D Burke RL Katzenstein D Association between human immunodeficiency virus and herpes simplex virus type 2 seropositivity among male factory workers in Zimbabwe J Infect Dis 1998 177 481 484 9466543
Obasi A Mosha F Quigley M Sekirassa Z Gibbs T Munguti K Todd J Grosskurth H Mayaud P Changalucha J Brown D Mabey D Hayes R Antibody to herpes simplex virus type 2 as a marker of sexual risk behavior in rural Tanzania J Infect Dis 1999 179 16 24 9841817 10.1086/314555
Mbopi-Keou FX Gresenguet G Mayaud P Weiss HA Gopal R Matta M Paul JL Brown DW Hayes RJ Mabey DC Belec L Interactions between herpes simplex virus type 2 and human immunodeficiency virus type 1 infection in African women: opportunities for intervention J Infect Dis 2000 182 1090 1096 10979904 10.1086/315836
Auvert B Ballard R Campbell C Carael M Carton M Fehler G Gouws E MacPhail C Taljaard D Van Dam J Williams B HIV infection among youth in a South African mining town is associated with herpes simplex virus-2 seropositivity and sexual behaviour Aids 2001 15 885 898 11399961 10.1097/00002030-200105040-00009
Shaw M van der Sande M West B Paine K Ceesay S Bailey R Walraven G Morison L McAdam K Prevalence of herpes simplex type 2 and syphilis serology among young adults in a rural Gambian community Sex Transm Infect 2001 77 358 365 11588283 10.1136/sti.77.5.358
Halton K Ratcliffe AA Morison L West B Shaw M Bailey R Walraven G Herpes simplex 2 risk among women in a polygynous setting in rural West Africa Aids 2003 17 97 103 12478074 10.1097/00002030-200301030-00013
Mertz GJ Schmidt O Jourden JL Guinan ME Remington ML Fahnlander A Winter C Holmes KK Corey L Frequency of acquisition of first-episode genital infection with herpes simplex virus from symptomatic and asymptomatic source contacts Sex Transm Dis 1985 12 33 39 2988143
Koutsky LA Stevens CE Holmes KK Ashley RL Kiviat NB Critchlow CW Corey L Underdiagnosis of genital herpes by current clinical and viral-isolation procedures N Engl J Med 1992 326 1533 1539 1315930
Baker DA Blythe JG Miller JM Once-daily valacyclovir hydrochloride for suppression of recurrent genital herpes Obstet Gynecol 1999 94 103 106 10389727 10.1016/S0029-7844(99)00239-2
Tyring SK Douglas JMJ Corey L Spruance SL Esmann J A randomized, placebo-controlled comparison of oral valacyclovir and acyclovir in immunocompetent patients with recurrent genital herpes infections. The Valaciclovir International Study Group Arch Dermatol 1998 134 185 191 9487210 10.1001/archderm.134.2.185
Corey L Wald A Patel R Sacks SL Tyring SK Warren T Douglas JMJ Paavonen J Morrow RA Beutner KR Stratchounsky LS Mertz G Keene ON Watson HA Tait D Vargas-Cortes M Once-daily valacyclovir to reduce the risk of transmission of genital herpes N Engl J Med 2004 350 11 20 14702423 10.1056/NEJMoa035144
Wald A Zeh J Barnum G Davis LG Corey L Suppression of subclinical shedding of herpes simplex virus type 2 with acyclovir Ann Intern Med 1996 124 8 15 7503497
Tyring SK Baker D Snowden W Valacyclovir for herpes simplex virus infection: long-term safety and sustained efficacy after 20 years' experience with acyclovir J Infect Dis 2002 186 Suppl 1 S40 6 12353186 10.1086/342966
Scoular A Gillespie G Carman WF Polymerase chain reaction for diagnosis of genital herpes in a genitourinary medicine clinic Sex Transm Infect 2002 78 21 25 11872854 10.1136/sti.78.1.21
Afonina I Zivarts M Kutyavin I Lukhtanov E Gamper H Meyer RB Efficient priming of PCR with short oligonucleotides conjugated to a minor groove binder Nucleic Acids Res 1997 25 2657 2660 9185578 10.1093/nar/25.13.2657
Kutyavin IV Afonina IA Mills A Gorn VV Lukhtanov EA Belousov ES Singer MJ Walburger DK Lokhov SG Gall AA Dempcy R Reed MW Meyer RB Hedgpeth J 3'-minor groove binder-DNA probes increase sequence specificity at PCR extension temperatures Nucleic Acids Res 2000 28 655 661 10606668 10.1093/nar/28.2.655
Stevenson J Hymas W Hillyard D Effect of sequence polymorphisms on performance of two real-time PCR assays for detection of herpes simplex virus J Clin Microbiol 2005 43 2391 2398 15872272 10.1128/JCM.43.5.2391-2398.2005
Ashley RL Laboratory techniques in the diagnosis of herpes simplex infection Genitourin Med 1993 69 174 183 8392966
Burrows J Nitsche A Bayly B Walker E Higgins G Kok T Detection and subtyping of Herpes simplex virus in clinical samples by LightCycler PCR, enzyme immunoassay and cell culture BMC Microbiol 2002 2 12 12069697 10.1186/1471-2180-2-12
van Dyck E Buve A Weiss HA Glynn JR Brown DW De Deken B Parry J Hayes RJ Performance of commercially available enzyme immunoassays for detection of antibodies against herpes simplex virus type 2 in African populations J Clin Microbiol 2004 42 2961 2965 15243045 10.1128/JCM.42.7.2961-2965.2004
Morrow RA Friedrich D Krantz E Performance of the Focus and Kalon Enzyme-Linked Immunosorbent Assays for Antibodies to Herpes Simplex Virus Type 2 Glycoprotein G in Culture-Documented Cases of Genital Herpes J Clin Microbiol 2003 41 5212 5214 14605166 10.1128/JCM.41.11.5212-5214.2003
Aldea C Alvarez CP Folgueira L Delgado R Otero JR Rapid Detection of Herpes Simplex Virus DNA in Genital Ulcers by Real-Time PCR Using SYBR Green I Dye as the Detection Signal J Clin Microbiol 2002 40 1060 1062 11880439 10.1128/JCM.40.3.1060-1062.2002
Tsurumi T Maeno K Y. N Nucleotide sequence of the DNA polymerase gene of herpes simplex virus type 2 and comparison with the type 1 counterpart. Gene 1987 52 129 137 3038677 10.1016/0378-1119(87)90039-4
Ye S Dhillon S Ke X Collins AR Day IN An efficient procedure for genotyping single nucleotide polymorphisms Nucleic Acids Res 2001 29 E88 8 11522844 10.1093/nar/29.17.e88
Ramaswamy M McDonald C Smith M Thomas D Maxwell S Tenant-Flowers M Geretti AM Diagnosis of genital herpes by real time PCR in routine clinical practice Sex Transm Infect 2004 80 406 410 15459412 10.1136/sti.2003.008201
Koutsky LA Ashley RL Holmes KK Stevens CE Critchlow CW Kiviat N Lipinski CM Wolner-Hanssen P L. C The frequency of unrecognized type 2 herpes simplex virus infection among women. Implications for the control of genital herpes. Sex Transm Dis 1990 17 90 94 2163115
Liew M Pryor R Palais R Meadows C Erali M Lyon E Wittwer C Genotyping of single-nucleotide polymorphisms by high-resolution melting of small amplicons Clin Chem 2004 50 1156 1164 15229148 10.1373/clinchem.2004.032136
|
16095535
|
PMC1236615
|
CC BY
|
2021-01-04 16:38:58
|
no
|
Virol J. 2005 Aug 11; 2:61
|
utf-8
|
Virol J
| 2,005 |
10.1186/1743-422X-2-61
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1618779510.1371/journal.pbio.0030380Research ArticleAnimal BehaviorEvolutionZoologyPrimatesFirst Observation of Tool Use in Wild Gorillas Tool Use in Wild GorillasBreuer Thomas [email protected]
1
2
Ndoundou-Hockemba Mireille
1
Fishlock Vicki
1
1 Wildlife Conservation Society, Mbeli Bai Study, Nouabalé-Ndoki Project, Brazzaville, Republic of Congo,2 Max Planck Institute for Evolutionary Anthropology, Leipzig, Germanyde Waal Frans Academic EditorEmory UniversityUnited States of America11 2005 1 10 2005 1 10 2005 3 11 e3802 2 2005 9 9 2005 Copyright: © 2005 Breuer 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.
Wild Gorillas Handy with a Stick
Descriptions of novel tool use by great apes in response to different circumstances aids us in understanding the factors favoring the evolution of tool use in humans. This paper documents what we believe to be the first two observations of tool use in wild western gorillas (Gorilla gorilla). We first observed an adult female gorilla using a branch as a walking stick to test water deepness and to aid in her attempt to cross a pool of water at Mbeli Bai, a swampy forest clearing in northern Congo. In the second case we saw another adult female using a detached trunk from a small shrub as a stabilizer during food processing. She then used the trunk as a self-made bridge to cross a deep patch of swamp. In contrast to information from other great apes, which mostly show tool use in the context of food extraction, our observations show that in gorillas other factors such as habitat type can stimulate the use of tools.
The first documentation of tool use by wild western gorillas
==== Body
Introduction
Tool use is defined as “the employment of an unattached environmental object to alter more efficiently the form, position, or condition of another object, another organism, or the user itself when the user holds or carries the tool during or just prior to use and is responsible for the proper and effective orientation of the tool” (p. 10 of [1]). Beck's classic book [1] defines six different types of tools: objects thrown at predators or rivals, objects used to hit predators, hunting weapons (only hominids), objects incorporated into social displays, objects to clean body parts, and objects made and used to acquire food, such as insects or nuts [1,2].
Information on tool use and factors favoring tool use in wild apes helps us to understand its importance in the evolution of our own species. Although there are reports of tool use by captive gorillas (Gorilla sp.), including object throwing and use of tools in feeding [3–9], there has been to our knowledge no reported case of tool use in by wild gorillas, despite decades of field research. It has been argued that gorilla tool use in captivity is less extensive than that shown by other apes [10], but it has recently been demonstrated that captive western gorillas (Gorilla gorilla) showed tool-using skills similar to those of orangutans (Pongo pygmaeus) [11], who frequently use tools in the wild [12].
One possible explanation for the absence of observed tool use in wild gorillas is that they are less dependent on extractive foraging techniques that might require the use of tools, since they exploit food resources differently than chimpanzees (Pan troglodytes) [2,13]. Whereas chimpanzee feeding ecology involves tools such as hammers to crack open nuts and sticks to fish for termites, gorillas access these food resources by breaking nuts with their teeth and smashing termite mounds with their hands. Nevertheless, mountain gorillas (Gorilla beringei) possess food-processing skills of comparable complexity and logical organization to chimpanzee termite fishing [14], which give them access to additional herbs in their habitat [15].
Here we report two instances of novel tool use by gorillas, both involving using detached branches to give postural support to cross water and swamp at Mbeli Bai, a forest clearing in northern Congo.
Results
On 9 October 2004, we observed George's group (monitored since 1995) near the edge of the clearing 180 m from the observation platform. This group had not been observed to use this area for more than 6 mo, and meanwhile elephants had created a new pool to feed on soil. Adult female Leah was seen at the pool edge near where a branch was sticking out of the surface, looking intently at the water in front of her for 1 min before she entering the water (Figure 1, top left). She began to cross the pool walking bipedally, but after her first steps the water quickly became waist deep and she returned to the pool edge. Leah then re-entered again bipedally and grabbed the straight branch in front of her with her right hand (Figure 1, top center). Relative to Leah's body size we estimated the leafless branch as being approximately 1 m long and 2 cm thick. Leah then detached the stick and, stretching forward with it in her right hand, seemed to use it to test the water depth or substrate stability: she grasped the stick firmly and repeatedly prodded the water in front of her with the end of the stick (Figure 1, top right). She then moved further into the pool, holding the detached branch in her right hand and using it as a walking stick for postural support (Figure 1, bottom three images). She advanced a further 8–10 m from the pool edge (not shown), repeating the actions shown in Figure 1, and then, leaving the stick in the pond, returned to her entry point, where her offspring was crying. She did not re-enter the water, but instead moved around the pool to feed on aquatic herbs.
Figure 1 Female Leah Using a Walking Stick while Crossing Bipedally through an Elephant Pool at Mbeli Bai
Female Leah first looked at the new elephant pool and the branch she later used as the walking stick, and entered the water without the tool (not shown). After re-entering the pool and taking the branch with her right hand, she walked bipedally 8–10 m into the water, frequently testing water deepness.
The second observation involved female Efi from Zulu group (monitored since 2000), who entered the clearing on 21 November 2004, 150 m from the observation platform. While close to the forest edge Efi detached a 1.3-m-long and 5-cm-thick leafless trunk of a dead shrub with both hands. She forcefully pushed it into the ground with both hands (Figure 2, middle) and held the tool for support with her left hand over her head for 2 min while dredging food with the other hand (Figure 2, bottom). Efi then took the trunk with both hands and placed it on the swampy ground in front of her, crossed bipedally on this self-made bridge, and walked quadrupedally towards the middle of the clearing (not shown). We could not see if the bridge was long enough to cross the swampy ground, but it certainly gave the female more stability underfoot and supported her weight for at least part of the distance.
Figure 2 Female Efi Using Trunk as a Stabilizer during Food Processing at Mbeli Bai
The top photo shows the intact trunk shortly before Efi manipulated it (visible to the left of female Fulani). The trunk was then detached by female Efi with both hands (middle), pushed into the ground, and used as a stabilizing stick while dredging aquatic herbs towards her with her other hand (bottom).
Discussion
It has been proposed that animals that can obtain resources by force may not need to rely on tools [2,13], but our observations show that other functional demands may also stimulate tool use, at least in apes. Our observations of gorilla tool use for postural support in a swampy habitat supports arguments that the use of tools reflects ecological needs and may therefore be viewed as an adaptation to a particular suite of environmental conditions [14]. Swamp habitat demands particular behavioral adaptations for western gorillas and may stimulate special behaviors, such as the splash display of adult male silverbacks at Mbeli Bai [16]. The forest edges of Mbeli Bai are heavily inundated, and gorillas often use branches to haul themselves out of the water or move around the edge of the clearing, but the cases reported here are the first observations, to our knowledge, of wild gorillas using detached objects as tools.
The observed tool use involved gorillas from two different groups and thus could indicate independent inventions, perhaps reflecting past negative experiences with deep water. However, there is also the intriguing possibility that using branches to test water depth or as bridges is a more common adaptation to this particular habitat. Footprints of gorillas walking over branches as bridges have been observed in other swampy clearings not far from Mbeli Bai [17].
We make no hypothesis about the mode of acquisition of this tool use behavior, but simply note that the high population density of western gorillas in northern Congo [18–20] makes social learning a possible transmission route [2].
All great apes use tools in captivity, but until recently tool use in the wild was only regularly observed in chimpanzees and orangutans [12,21,22]. There are now reports of tool use by bonobos (P. paniscus) [23] and western gorillas, which highlights the importance of long-term studies in these species.
Materials and Methods
Mbeli Bai is a 12.9-ha large swampy forest clearing in the in the southwest part of Nouabalé-Ndoki National Park, Republic of Congo. The clearing includes streams and elephant pools and is dominated by aquatic vegetation in the Cyperaceae, Gramineae, and Hydrocharitaceae families. The aquatic vegetation forms a floating surface to the bai, with an average depth of 1 m [24]. Monitoring of the social organization, demography, and behavior of western gorillas at Mbeli Bai has been ongoing since 1995 [25,26]. Observations are made with 15–45 × 60 mm telescopes from a 9-m-high viewing platform overlooking the clearing. Gorillas are habituated to the presence of researchers and can be identified by their distinctive features, such as pelage, nose prints, and shape of brow ridges [25].
We thank the Ministère de l'Economie Forestière and the Nouabalé-Ndoki Project of the Wildlife Conservation Society for permission to work in Nouabalé-Ndoki National Park. We further thank the staff of the Wildlife Conservation Society Congo Program and the Nouabalé-Ndoki Project for logistical and administrative support. Financial support for the Mbeli Bai Study was provided by the Columbus Zoo and Aquarium, Cincinnati Zoo and Botanical Garden, Sea World and Busch Gardens Conservation Fund, Woodland Park Zoo, Toronto Zoo, and Max Planck Society. We thank Crickette Sanz for helping to improve the final version of the manuscript.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. TB conceived and designed the study. TB, MNH, and VF performed the study. TB analyzed the data and contributed reagents/materials/analysis tools. TB and VF wrote the paper.
Citation: Breuer T, Ndoundou-Hockemba M, Fishlock V (2005) First observation of tool use in wild gorillas. PLoS Biol 3(11): e380.
==== Refs
References
Beck BB Animal tool behavior: The use and manufacture of tools by animals 1980 New York Garland Press 307
Van Schaik CP Deaner RO Merrill MY The conditions for tool use in primates: Implications for the evolution of material culture J Hum Evol 1999 36 719 741 10330335
Boysen ST Kuhlmeier VA Halliday PO Halliday YM Parker ST Mitchell RW Miles HL Tool use in captive gorillas The mentalities of gorillas and orangutans: Comparative perspectives 1999 Cambridge Cambridge University Press 179 187
Wood RJ Spontaneous use of sticks by gorillas at Howlett's Zoo Park, England Int Zoo News 1984 31 13
Natale P Poti' P Spinozzi G Development of tool use in a macaque and a gorilla Primates 1988 29 413 416
Gomez JC Tool-use and communication as alternative strategies of problem-solving in the gorilla Primate Rep 1988 19 25 28
Nakamichi M Stick throwing by gorillas (Gorilla gorilla gorilla) at the San Diego Wild Animal Park Folia Primatol 1998 69 291 295 9751834
Nakamichi M Spontaneous use of sticks as tools by captive gorillas (Gorilla gorilla gorilla)
Primates 1999 40 487 498
Fontain B Moisson PY Wickings EJ Observations of spontaneous tool making and tool use in a captive group of western lowland gorillas (Gorilla gorilla gorilla)
Folia Primatol 1995 65 219 223
Tomasello M Call J Primate cognition 1997 New York Oxford University Press 528
Mulcahy N Call J Dunbar RIM Gorillas (Gorilla gorilla) and orangutans (Pongo pygmaeus) encode relevant problem features in a tool-using task J Comp Psychol 2005 119 23 32 15740427
Fox EA Sitompul AF van Schaik CP Parker ST Mitchell RW Miles HL Intelligent tool use in wild Sumatran orangutans The mentalities of gorillas and orangutans: Comparative perspectives 1999 Cambridge Cambridge University Press 99 116
McGrew WC Why is ape tool use so confusing? In: Standen V, Foley RA, editors. Comparative socioecology: The behavioural ecology of humans and other mammals 1989 Oxford Blackwell Scientific Publications 457 472
Byrne R The thinking ape: Evolutionary origins of intelligence 1995 Oxford Oxford University Press 266
Byrne RW Byrne JME Complex leaf-gathering skills of mountain gorillas (Gorilla g. beringei) Variability and standardization Am J Primatol 1993 31 241 261
Parnell RJ Buchanan-Smith H An unusual display by gorillas Nature 2001 412 294 11460152
Dzomambou SA Nishihira T Rapport sur l'etablissement d'un pond dans une saline par les gorilles. Recherche scientifique cooperatives par les equipes Japonaises et Congolais Annual report to the Congolese government (Ministaire de l'Economie Forestière) 1996
Blake S, Rogers E, Fay JM, Ngangoué M, Ebéké Swamp gorillas in northern Congo Afr J Ecol 1995 33 285 290
Poulson JR Clark CJ Densities, distributions, and seasonal movements of gorillas and chimpanzees in swamp forests in northern Congo Int J Primatol 2004 25 285 306
Morgan D, Sanz C, Onononga JR, Strindberg S Ape abundance and habitat use in the Goualougo Triangle, Republic of Congo Am J Primatol 2005 In press
McGrew WC Chimpanzee material culture: Implications for human evolution 1992 Cambridge Cambridge Universtiy Press 277
McGrew WC The cultured chimpanzee 2004 Cambridge Cambridge University Press 244
Hohmann G Fruth B Culture in bonobos? Between-species and within-species variation in behaviour Curr Anthropol 2003 44 563 571
Parnell RJ The social structure and behaviour of western lowland gorillas (Gorilla gorilla gorilla) at Mbeli Bai, Republic of Congo [dissertation] 2002 Stirling (United Kingdom) University of Stirling 340
Parnell RJ Group size and structure in western lowland gorillas (Gorilla gorilla gorilla) at Mbeli Bai, Republic of Congo Am J Primatol 2002 56 193 206 11948636
Stokes EJ Parnell RJ Olejniczak C Female dispersal and reproductive success in wild western lowland gorillas (Gorilla gorilla gorilla)
Behav Ecol Sociobiol 2003 54 329 339
|
16187795
|
PMC1236726
|
CC BY
|
2021-01-05 08:21:27
|
no
|
PLoS Biol. 2005 Nov 1; 3(11):e380
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030380
|
oa_comm
|
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030385SynopsisAnimal BehaviorEvolutionZoologyPrimatesWild Gorillas Handy with a Stick Synopsis11 2005 1 10 2005 1 10 2005 3 11 e385Copyright: © 2005 Public Library of Science.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.
First Observation of Tool Use in Wild Gorillas
==== Body
When chimpanzees were first seen using tools in Liberia in 1951, little was known about great ape behavior in the wild. The sighting was published as a short note by Harry Beatty in the Journal of Mammalogy (recounted by primatologist Frans de Waal in The Ape and the Sushi Master). After spotting a shell mound, Beatty saw a chimp “come ambling round a bend bearing an armful of dried palm nuts,” sit down beside a rock, select a nut, place it on a flat rock surface, and pound it with another rock to extract the meat. Thirty years passed before primatologists would describe the same behavior in chimps in Guinea. By then it was clear that nonhuman primates used objects for many tasks, from intimidating rivals and predators to taking leaf sponge baths and processing food. Still, the notion that tool use was not the sole province of human intelligence was difficult for some to accept.
Though chimps, bonobos, orangutans, and gorillas all use tools in captivity, often imitating behaviors of their keepers, tool use had not been observed for gorillas in the wild. (Chimps famously use sticks to fish for termites and rocks to crack nuts in the wild; orangutans use sticks to extract seeds from the spiny, sharp husk of Neesia fruits in the wild, and bonobos use various tools, including moss as sponges and leafy twigs to swat away bees.) That no one has ever observed wild gorillas using tools—despite decades of intensive study—prompted scientists to speculate that gorillas lost the skill because they didn't need it. The largest of the great apes, gorillas can gnash nuts between their teeth and simply smash termite mounds to release their denizens.
A first time for everything: This adult female gorilla in Nouabalé-Ndoki National Park, northern Congo, uses a branch as a walking stick to gauge the water's depth, proving that gorillas use tools too
But now, as part of an ongoing study of western gorillas in Nouabalé-Ndoki National Park in the Republic of Congo, Thomas Breuer, Mireille Ndoundou-Hockemba, and Vicki Fishlock reveal that gorillas are just as resourceful as the other great apes. From an observation platform at Mbeli Bai, a swampy forest clearing that gorillas frequently visit to forage, Breuer et al. observed an adult female gorilla named Leah (a member of a long-studied gorilla group) at the edge of a pool of water, “looking intently at the water in front of her.” Leah walked upright into the water, but stopped and returned to the edge when the water reached her waist. She then walked back into the water, grabbed a branch in front of her, detached it, and, grasping it firmly, repeatedly jabbed the water in front of her with the end of the branch, “apparently using it to test the water depth or substrate stability.” She continued walking across the pool, branch in hand, “using it as a walking stick for postural support.”
In a separate incident, a second adult female named Efi (who belonged to a different group) emerged from the forest and broke off the thin, long trunk from a dead shrub with both hands. She then jammed the trunk into the ground and grasped it with her left hand while manipulating food with her right hand. Efi then placed the trunk on the swampy ground in front of her, and, walking upright, used it as a bridge to cross the muck.
That tool use was seen in lowland gorillas and not mountain gorillas—the most intensively studied gorillas—may reflect the different ecological challenges facing the two populations, supporting the notion that ecology influences the evolution of novel tool use. Though Mbeli Bai gorillas often use intact branches to help them get around, the authors explain, these are the first observations of wild gorillas using detached objects as tools. However the gorillas acquired their skills, these observations suggest that the intelligence required for tool use evolved before the gorilla lineage split off from humans and the other great apes—providing further evidence that intelligence is not unique to humans. —Liza Gross
|
0
|
PMC1236727
|
CC BY
|
2021-01-05 08:21:29
|
no
|
PLoS Biol. 2005 Nov 1; 3(11):e385
|
utf-8
|
PLoS Biol
| 2,005 |
10.1371/journal.pbio.0030385
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383110.1371/journal.pmed.0020266Research in TranslationGenetics/Genomics/Gene TherapyImmunologyOtherPharmacology/Drug DiscoveryHealth PolicyRespiratory MedicinePsychiatrySubstance use (including alcohol)Respiratory MedicineSmokingPublic HealthWill Nicotine Genetics and a Nicotine Vaccine Prevent Cigarette Smoking and Smoking-Related Diseases? Research in TranslationHall Wayne D Wayne D. Hall is at the Office of Public Policy and Ethics, Institute for Molecular Bioscience, University of Queensland, St. Lucia, Queensland, Australia. E-mail: [email protected]
Competing Interests: The work reported in this paper was funded by start-up funding for the Office of Public Policy and Ethics provided by the Institute for Molecular Bioscience and the Vice Chancellor's Strategic Fund of the University of Queensland.
9 2005 27 9 2005 2 9 e266Copyright: © 2005 Wayne D. Hall.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.
Test Your Knowledge: Ten Questions on Tobacco Addiction
Hall argues that the preventive use of genetic and vaccine biotechnologies is a superficially attractive tobacco policy option of doubtful efficacy, cost-effectiveness, and ethicality.
==== Body
Biotechnologies that aim to prevent smoking and tobacco-related disease may emerge as unintended by-products of research on the genetics of nicotine dependence and the effectiveness of nicotine vaccines for smoking cessation. For many advocates of tobacco control, any discussion of the genetics of smoking is anathema because of the tobacco industry's use of Fisher's [1] genetic hypothesis to argue that cigarette smoking did not cause lung cancer [2,3]. Advocates of tobacco control say that the goal of tobacco policy should be to eliminate cigarette smoking by imposing high taxes on tobacco products, preventing the tobacco industry from promoting its products, restricting access to tobacco and opportunities to smoke, and minimising nonsmokers' exposure to environmental tobacco smoke [4]. These policies have substantially reduced smoking prevalence in the developed countries that have adopted them [4].
Even if one agrees with these policies, as I do, it would be unwise to ignore evidence that genetic factors play a role in smoking initiation and persistence or neurobiological explanations of why some smokers find it so difficult to quit. Indeed, the past misuse of genetic research on nicotine and the possible public misunderstanding of the role of nicotine vaccines make it all the more important for public health practitioners and genetic and neurobiological researchers to be well acquainted with some of the superficially attractive but mistaken ways in which this work may be used.
In this short paper, I briefly explain why it would be unwise to use genetic and neurobiological knowledge to prevent cigarette smoking and tobacco-related disease. However implausible these uses may seem to those who are well informed about the genetics of tobacco use or tobacco-control policy, it is the preventive uses of genetic information and nicotine vaccines that most excite the interest of the media and the public. The major challenges that these approaches face need to be widely understood if we are to prevent these superficially attractive but controversial uses from undermining effective control policies and the development of better methods of helping smokers to quit.
The Genetics of Smoking
Twin studies of cigarette smoking [5,6] estimate that the heritability of smoking initiation is 50% and that for smoking persistence is 70% [5,7,8]. There are a number of plausible “candidate genes” that predict an increased risk of nicotine dependence [5,8]. These include polymorphisms that affect nicotine metabolism, as well as dopamine receptors and transporters that mediate reward in the nucleus accumbens [9]. The most plausible hypothesis is that multiple genes are involved in smoking initiation and persistence [5,7,10,11], each of which only modestly increases the risk of developing dependence. The relative risks for the alleles that have been most consistently associated with smoking initiation, adoption, persistence, and cessation are typically less than 1.30 [5], well within the range of associations observed between polymorphisms and other human diseases, namely, 1.2–1.5 [12].
Predictive Testing for Genetic Risks of Nicotine Dependence
The predictive testing of genetic risk dependence is one of the first uses that journalists often suggest for research on the genetics of nicotine dependence. In this scenario, the population would be screened for susceptibility alleles with the aim of giving preventive behavioural and pharmacological interventions to individuals who are at higher genetic risk of smoking [13]. There is an obvious objection to this proposal: that it is not good public health policy to encourage people to smoke tobacco, regardless of their genetic risk of dependence [7].
Figure 1 Gross Pathology of the Lung Showing Centrilobular Emphysema Characteristic of Smoking
(Photo: Edwin P. Ewing, Jr./Centers for Disease Control and Prevention)
An alternative rationale is that screening would allow individuals who were at highest genetic risk of nicotine dependence to make an informed decision to avoid tobacco smoking. Even if we value individual autonomy highly, genetic screening for nicotine dependence is unlikely to be a good policy for the following reasons [14].
First, when multiple genes predispose to a condition, individual susceptibility alleles only predict a very modestly increased risk of dependence [15]. Testing multiple alleles would improve prediction if the results of the multiple tests were combined [16,17], but the larger the number of genes that are involved in disease susceptibility, the less useful most individuals will find information about their genotype. This is because as the number of alleles increases, the risk distribution tends to the log normal [17], a distribution in which the number of persons with very high risk combinations of multiple genes is very small and in which the majority will be at average genetic risk [7,16].
Second, the above means that a very large number of individuals would need to be screened to identify the few at highest risk [18,19]. For example, Yang et al. [20] simulated screening for 5 susceptibility alleles, each with a relative risk (RR) ranging from 1.5 to 3.5 and a prevalence in the population between 0.10 and 0.25 (all within the range of empirical data). One of these alleles was assumed to interact with an environmental exposure with a prevalence of 15% and a RR of 2.0. Their simulation showed that those who screened positively on all five genes had an 81% chance of developing the disease; this rose to 89% if they had also been exposed to the environmental risk factor. However, these risks applied to only four persons in 100,000, and only 79 in 100,000 would have a greater than 50% risk.
Third, predictive genetic testing may have unintended adverse effects. This would be the case, for example, if testing adolescents for susceptibility to nicotine dependence increased their preparedness to initiate smoking—as could happen, for example, if they were prompted to test the accuracy of the genetic predictions by smoking [7].
Vaccination against nicotine could reduce relapse to smoking in abstinent smokers
(Photo: Bill Branson/National Cancer Institute)
Fourth, screening is only ethically justifiable if there is an effective intervention to prevent the disorder in those who are identified as being at increased genetic risk [15,21]. “Avoid smoking” is good health advice, regardless of one's genetic susceptibility. The prospect of combining genetic screening with some preventive intervention sounds a more appealing option, especially when such an intervention—a vaccine against nicotine—is being developed and trialled by three pharmaceutical companies for smoking cessation [22].
Preventive Use of a Nicotine Vaccine
A “nicotine vaccine” induces the immune system to produce antibodies that bind to nicotine and prevent it from crossing the blood–brain barrier to act on receptors in the brain [23–25]. Animal studies have shown that attaching nicotine to a suitable antigenic protein (e.g., [23,24]) produces antibodies that have a high affinity for nicotine [23,25]. Vaccination of animals attenuates nicotine effects [24], abolishes nicotine self-administration [23], and suppresses dopamine release in the nucleus accumbens [26].
Active vaccination against nicotine could reduce relapse to smoking in abstinent smokers by attenuating the pharmacological effects of nicotine during the first few months after quitting, when most smokers relapse [25]. A nicotine vaccine could be circumvented by increasing the dose of nicotine, but attenuating the rewarding effects of nicotine may nonetheless be enough to reduce relapse rates in smokers by making a lapse less likely to lead to a return to daily smoking [22,25]. This promising immunological technology for improving the success of smoking cessation is currently in trials [22,25].
The term “vaccine” inevitably prompts journalists to ask about its possible preventive use. Misconceptions that a vaccine will produce lifelong immunity against nicotine may prompt parents to vaccinate their children [27]. As minors, children would not be legally able to consent to vaccination, but since parents already make choices for their children about other vaccines and interventions that affect their lives (e.g., their diet and education), some have argued that vaccination against nicotine and other drugs is simply another decision that parents should be able to make on behalf of their children [27]. This argument is likely to be contested by civil libertarians and by adolescents who disagree with their parents' wishes. One can expect the tobacco industry to amplify such dissenting views.
Even if we set aside the ethical issues, there are major practical obstacles to the preventive use of a nicotine vaccine in children. First, the limited period of protection provided by existing vaccines would require booster injections, perhaps every two or three months throughout adolescence [28]. Second, the fact that the vaccine could be circumvented by using higher doses of nicotine means that vaccination could be counterproductive if adolescents were prompted to test its efficacy. Third, it would be costly to universally vaccinate children against nicotine with a vaccine that would probably have only modest preventive efficacy. These obstacles (and the high regulatory hurdles that such a vaccine might be expected to have to leap when used preventively in children) make it unlikely that universal nicotine vaccination would be publicly funded.
“Indicated vaccination” of “high risk” adolescents seems a more plausible option because it would be much less expensive to only vaccinate young people who are at “high genetic risk” of smoking. The feasibility of this approach is also doubtful, given the likely low predictive validity of genetic screening for smoking risk (outlined above), the doubtful preventive efficacy of a nicotine vaccine, and the possible adverse effects of vaccination, such as stigmatisation of those who screen positive, and discrimination against them by third parties, such as life or health insurance companies.
The “off-label” use of a nicotine vaccine by a physician acting at the request of a parent is the most likely way in which a vaccine will be used preventively. It is difficult to see how this can be prevented if a nicotine vaccine is approved for therapeutic use, other than by education of physicians and parents about the limitations of this approach.
Screening for Genetic Susceptibility to Tobacco-Related Diseases
Genetic factors also appear to play a role in susceptibility to many nonfamilial types of cancer, although there is disagreement about how large a role this is [29–32]. There is evidence that polymorphisms that affect the metabolism of carcinogens in tobacco smoke and repair of damage to DNA may increase the risk of smokers developing lung cancer [33–35]. There are also indications that polymorphisms may affect the likelihood of smokers developing heart disease [36] and chronic obstructive lung disease (Figure 1) [37].
Many ambivalent smokers may be attracted by the superficially plausible idea that they could continue to smoke with impunity if they did not have any of the alleles that predict an increased risk of smoking-related diseases [38]. However, this type of screening is very unlikely to work, for reasons that need to be widely understood by the community.
There are major practical obstacles to the preventive use of a nicotine vaccine in children.
First, if, as seems most likely, multiple genes are involved in susceptibility to multiple tobacco-related diseases, then the ability to predict tobacco-related disease risk from genetic tests may not improve on the prediction of disease risk from smoking status.
Second, cigarette smoking causes multiple diseases, with lung and other cancers, heart disease and chronic obstructive lung disease being the most prevalent. This means that predicting one's genetic risk of only the most common tobacco-related diseases would involve testing a large number of polymorphisms.
Third, because multiple susceptibility alleles would have to be tested for multiple diseases, most smokers would be at increased genetic risk of developing at least one smoking-related disease [38]. For example, if we screened for six different susceptibility alleles (each with a RR of 1.5, a prevalence of 10%, and multiplicative risks) for each of five major tobacco-related diseases (lung cancer, coronary heart disease, chronic lung disease, other cancers, and stroke), then only 3% of the screened population would not be at increased risk of developing any of the diseases. Conversely, 97% of smokers would be at increased risk of developing at least one of these diseases if they continued to smoke.
Conclusion
The preventive use of genetic and vaccine biotechnologies—screening the population for genetic susceptibility to nicotine dependence, vaccinating children who do not smoke against the effects of nicotine, and screening smokers for polymorphisms that predict increased susceptibility to tobacco-related diseases—are superficially attractive tobacco policy options that are of doubtful efficacy, cost-effectiveness, and ethicality. We must ensure that the speculative use of these technologies does not undermine effective tobacco-control policies and the development of more effective ways of helping cigarette smokers to quit.
I would like to thank Sarah Yeates for her invaluable assistance in locating the research literature and in preparing this paper for publication, and Kate Morley for her advice on the genetics of nicotine dependence.
Citation: Hall WD (2005) Will nicotine genetics and a nicotine vaccine prevent cigarette smoking and smoking-related diseases? PLoS Med 2(9): e266.
Abbreviation
RRrelative risk
==== Refs
References
Fisher RA Cigarettes, cancer and statistics Centen Rev 1958 2 151 166
Kozlowski LT Rehabilitating a genetic perspective in the study of tobacco and alcohol use Br J Addict 1991 86 517 520 1859914
Stolley PD When genius errs: R.A. Fisher and the lung cancer controversy Am J Epidemiol 1991 133 416 425 2000852
Rabin RL Sugarman SD Regulating tobacco 2001 Oxford Oxford University Press 299
Lerman C Berrettini W Elucidating the role of genetic factors in smoking behavior and nicotine dependence Am J Med Genet B Neuropsychiatr Genet 2003 118 48 54
True WR Heath AC Scherrer JF Waterman B Goldberg J Genetic and environmental contributions to smoking Addiction 1997 92 1277 1287 9489045
Hall WD Madden P Lynskey M The genetics of tobacco use: Methods, findings and policy implications Tob Control 2002 11 119 124 12035004
Munafo M Clark T Johnstone E Murphy M Walton R The genetic basis for smoking behavior: A systematic review and meta-analysis Nicotine Tob Res 2004 6 583 597 15370155
Munafo M Johnstone E Murphy M Walton R New directions in the genetic mechanisms underlying nicotine addiction Addict Biol 2001 6 109 117 11341850
Crabbe JC Genetic contributions to addiction Annu Rev Psychol 2002 53 435 462 11752492
Uhl GR Liu QR Naiman D Substance abuse vulnerability loci: Converging genome scanning data Trends Genet 2002 18 420 425 12142011
Ioannidis JPA Trikalinos TA Ntzani EE Contopoulos-Ioannidis DG Genetic associations in large versus small studies: An empirical assessment Lancet 2003 361 567 571 12598142
Collins FS Shattuck lecture—Medical and societal consequences of the Human Genome Project N Engl J Med 1999 341 28 37 10387940
Holtzman NA Marteau TM Will genetics revolutionize medicine? N Engl J Med 2000 343 141 144 10891526
Evans JP Skrzynia C Burke W The complexities of predictive genetic testing BMJ 2001 322 1052 1056 11325775
Khoury MJ Yang Q Gwinn M Little J Dana Flanders W An epidemiologic assessment of genomic profiling for measuring susceptibility to common diseases and targeting interventions Genet Med 2004 6 38 47 14726808
Pharoah PD Antoniou A Bobrow M Zimmern RL Easton DF Polygenic susceptibility to breast cancer and implications for prevention Nat Genet 2002 31 33 36 11984562
Holtzman NA Shapiro D Genetic testing and public policy BMJ 1998 316 852 856 9549463
Vineis P Schulte P McMichael AJ Misconceptions about the use of genetic tests in populations Lancet 2001 357 709 712 11247571
Yang Q Khoury MJ Botto L Friedman JM Flanders WD Improving the prediction of complex diseases by testing for multiple disease-susceptibility genes Am J Med Genet 2003 72 636 649
Khoury MJ McCabe LL McCabe ER Population screening in the age of genomic medicine N Engl J Med 2003 348 50 58 12510043
Hall WD The prospects for immunotherapy in smoking cessation Lancet 2002 360 1089 1091 12384004
Lindblom N de Villiers SH Kalayanov G Gordon S Johansson AM Active immunization against nicotine prevents reinstatement of nicotine-seeking behavior in rats Respiration 2002 69 254 260 12097770
Pentel P Malin D A vaccine for nicotine dependence: Targeting the drug rather than the brain Respiration 2002 69 193 197 12097758
Vocci FJ Chiang CN Vaccines against nicotine: How effective are they likely to be in preventing smoking? CNS Drugs 2001 15 505 514 11510621
de Villiers SH Lindblom N Kalayanov G Gordon S Malmerfelt A Active immunization against nicotine suppresses nicotine-induced dopamine release in the rat nucleus accumbens shell Respiration 2002 69 247 253 12097769
Cohen PJ Immunization for prevention and treatment of cocaine abuse: Legal and ethical implications Drug Alcohol Depend 1997 48 167 174 9449015
Kosten TR Rosen M Bond J Settles M Roberts JS Human therapeutic cocaine vaccine: Safety and immunogenicity Vaccine 2002 20 1196 1204 11803082
Czene K Lichtenstein P Hemminki K Environmental and heritable causes of cancer among 9.6 million individuals in the Swedish Family-Cancer Database Int J Cancer 2002 99 260 266 11979442
Hemminki K Lonnstedt I Vaittinen P Lichtenstein P Estimation of genetic and environmental components in colorectal and lung cancer and melanoma Genet Epidemiol 2001 20 107 116 11119300
Lichtenstein P Holm NV Verkasalo PK Iliadou A Kaprio J Environmental and heritable factors in the causation of cancer—Analyses of cohorts of twins from Sweden, Denmark, and Finland N Engl J Med 2000 343 78 85 10891514
Risch N The genetic epidemiology of cancer: Interpreting family and twin studies and their implications for molecular genetic approaches Cancer Epidemiol Biomarkers Prev 2001 10 733 741 11440958
Bouchardy C Benhamou S Jourenkova N Dayer P Hirvonen A Metabolic genetic polymorphisms and susceptibility to lung cancer Lung Cancer 2001 32 109 112 11325480
Nasca PC Nasca PC Pestides H Tobacco and cancer Fundamentals of cancer epidemiology 2001 Gaithersburg (Maryland) Aspen 139 174
Wei Q Spitz MR The role of DNA repair capacity in susceptibility to lung cancer: A review Cancer Metastasis Rev 1997 16 295 307 9433641
Humphries SE Talmud PJ Hawe E Bolla M Day IN Apolipoprotein E4 and coronary heart disease in middle-aged men who smoke: A prospective study Lancet 2001 358 115 119 11463413
Lomas DA Silverman EK The genetics of chronic obstructive pulmonary disease Respir Res 2001 2 20 26 11686861
Wang XL Mahaney MC Genotype-specific effects of smoking on risk of CHD Lancet 2001 358 87 88 11463406
|
16173831
|
PMC1236791
|
CC BY
|
2021-01-05 10:40:31
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e266
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020266
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383210.1371/journal.pmed.0020269Correspondence and Other CommunicationsOncologyRespiratory MedicineAcquired Gefitinib-Resistant Mutation of EGFR in a Chemonaive Lung Adenocarcinoma Harboring Gefitinib-Sensitive Mutation L858R CorrespondenceGow Chien-Hung
1
Shih Jin-Yuan
1
Chang Yih-Leong
1
Yu Chong-Jen
1
1National Taiwan University HospitalTaipeiTaiwanE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
9 2005 27 9 2005 2 9 e269Copyright: © 2005 Gow 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.
Acquired Resistance of Lung Adenocarcinomas to Gefitinib or Erlotinib Is Associated with a Second Mutation in the EGFR Kinase Domain
==== Body
The research article by Pao et al. [1] provides important new information addressing three patients with acquired resistance to gefitinib or erlotinib in progressing tumors containing a secondary mutation, leading to the substitution of methionine for threonine at position 790 (T790M) in exon 20. However, all of the patients received systemic chemotherapy prior to gefitinib or erlotinib therapy, and the original lung tissue was obtained long before epidermal growth factor receptor (EGFR) inhibitors were used. We describe a chemonaive patient with gefitinib-sensitive lung adenocarcinoma harboring L858R. The tumor progressed and developed an additional T790M mutation after nine months of gefitinib treatment.
A 56-year-old female who had never smoked presented with nonproductive cough for one month. Her chest radiography revealed a mass in the right lower lung (RLL) (Figure 1A). Chest tomography (CT) confirmed a mass with pleural invasion and multiple small nodules in the bilateral lungs. Ultrasound-guided percutaneous transthoracic lung biopsy revealed adenocarcinoma. Gefitinib (250 mg/day) was initiated. The RLL tumor decreased in size significantly two months after treatment (Figure 1B). Both serum CEA and CA-199 decreased, from 4,178 ng/ml to 9.1 ng/ml and from 464 U/ml to 22 U/ml, respectively. However, the patient could not tolerate the severe side effects, including diarrhea, erythematous papules over the nasolabial areas and buttocks, and paronychia with granulation on fingers. Gefitinib was withdrawn for two weeks. Then, she received gefitinib at 250 mg on alternate days. These side effects became tolerable, and gefitinib at 250 mg/day was resumed. Nine months after initiating gefitinib, chest radiography revealed progression of tumor (Figure 1C). At this time, chest CT revealed tumor progression with an endobronchial tumor in the right middle bronchus. Gefitinib was discontinued. After obtaining written, informed consent from the patient, a CT-guided lung biopsy specimen was obtained. Pathological analysis confirmed the presence of adenocarcinoma. This patient received subsequent chemotherapy for advanced lung cancer.
Figure 1 Chest Radiography
Chest radiography shows a large mass in the RLL before gefitinib treatment (A), and marked decrease in tumor size two months after gefitinib was initiated (B). This tumor progressed nine months after gefitinib treatment (C).
Genomic DNA was extracted from the tumor specimen of an original lung biopsy and a progressive tumor biopsy specimen. The tyrosine kinase domain (exons 18–21) was amplified and sequenced. Mutations were also checked against the corresponding DNA from blood lymphocytes at the diagnosis of lung cancer. The original diagnostic biopsy specimen contained a thymidine to guanine mutation at nucleotide 2573 of exon 21, resulting in L858R. In the second biopsy, an additional single-base change from cytosine to thymidine was identified at nucleotide 2369 in exon 20, resulting in T790M.
This report strengthens the evidence of T790M as an acquired gefitinib-resistant mutation. Gefitinib responsiveness results in large part from the drug's effective inhibition of essential antiapoptotic signals transduced by the mutant receptor, and L858R is the most commonly detected mutation [2–5]. The T790M mutation is rarely found in tumors from patients not treated with either gefitinib or erlotinib, and could be discovered only in progressing tumors, in addition to a primary drug-sensitive mutation in EGFR. A non-small-cell lung cancer cell line bearing both T790M and L858R mutations was approximately 100-fold less sensitive to gefitinib or erlotinib, and did not show inhibition of tyrosine phosphorylation in vitro [1].
Pao et al. and Kobayashi et al. identified four cases with lung adenocarcinoma harboring pre-existing mutations of EGFR as delL747–E749 plus A750P, delE746–A750, or delL747–S752, prior to the use of gefitinib or erlotinib [1,6]. All of them had exposure to previous systemic chemotherapies and took a small-molecule tyrosine kinase inhibitor as the second- or third-line therapy, then all acquired a second mutation T790M in the following months after disease progression. In the case of our patient, the patient received no prior systemic chemotherapy, and we identified an initial gefitinib-sensitizing L858R EGFR mutation, followed by a T790M mutation concomitantly with L858R in the biopsy taken from the growing tumor nine months after gefitinib use. Though it is unlikely that prior chemotherapy led to the development of T790M mutation, given the complexity of EGFR mutation, further studies are still required to elucidate the role of T790M mutation in the context of EGFR mutations.
This correspondence was peer reviewed.
Citation: Gow CH, Shih JY, Chang YL, Yu CJ (2005) Acquired gefitinib-resistant mutation of epidermal growth factor receptor in a chemonaive lung adenocarcinoma harboring gefitinib-sensitive mutation L858R. PLoS Med 2(9): e269.
==== Refs
References
Pao W Miller VA Politi KA Riely GJ Somwar R Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain PLoS Med 2005 2 e73 10.1371/journal.pmed.0020073 15737014
Sordella R Bell DW Haber DA Settleman J Gefitinib-sensitizing EGFR mutations in lung cancer activate anti-apoptotic pathways Science 2004 305 1163 1167 15284455
Lynch TJ Bell DW Sordella R Gurubhagavatula S Okimoto RA Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib N Engl J Med 2004 350 2129 2139 15118073
Paez JG Janne PA Lee JC Tracy S Greulich H EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy Science 2004 304 1497 1500 15118125
Pao W Miller V Zakowski M Doherty J Politi K EGF receptor gene mutations are common in lung cancers from “never smokers” and are associated with sensitivity of tumors to gefitinib and erlotinib Proc Natl Acad Sci U S A 2004 101 13306 13311 15329413
Kobayashi S Boggon TJ Dayaram T Janne PA Kocher O EGFR mutation and resistance of non-small-cell lung cancer to gefitinib N Engl J Med 2005 352 786 792 15728811
|
16173832
|
PMC1236792
|
CC BY
|
2021-01-05 10:40:31
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e269
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020269
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383310.1371/journal.pmed.0020279Correspondence and Other CommunicationsPsychiatrySchizophrenia and Other Psychotic DisordersSecondary Schizophrenia CorrespondenceHambidge Dave
1
1StaffordshireUnited KingdomE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
9 2005 27 9 2005 2 9 e279Copyright: © 2005 Dave Hambidge.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 Systematic Review of the Prevalence of Schizophrenia
Authors' Reply: Measurement Errors in Schizophrenia Epidemiology
==== Body
May I respectfully highlight a potential confounding factor to interpreting an otherwise excellent and provoking study by Saha et al. [1].
In their recent overview of secondary schizophrenia (defined as “a disparate range of brain disorders that can, uncommonly, give rise to schizophrenia like symptomology” [2]), Hyde and Lewis [2] concluded that, overall, there was a prevalence rate of 5%–8% for psychoses of likely identifiable organic etiology amongst a series of relatively unselected patients. They suggest screening procedures in new cases of psychosis, including schizophrenia, with a battery of blood tests, a urine drug screen (UDS), and an electroencephalogram (EEG) as first-line investigations.
Between September 2000 and November 2003, I interviewed and studied the medical records of 56 patients in northwest England, who were appealing against detention under the Mental Health Act (1983), and who had been admitted for the first time within the last ten years [3]. They were all referred to me by their solicitors to prepare Legal Aid/Legal Services Commission–funded independent reports for their Mental Health Review Tribunal hearings. For each patient, I recorded which of the organic investigations suggested by Hyde and Lewis, if any, had been undertaken.
Of the 56 patients, ten were being detained for the first time (three females and seven males, detained on average for 39 weeks) and 13 had been detained for over one year (two females and 11 males, detained on average for 106 weeks). Whilst all except two of the 56 patients had some combination of blood tests recorded, 55% did not have a UDS and 83% did not have an EEG. Syphilis serology was examined for in only two patients of the latter group and none of the former. Therefore, my findings suggest that secondary schizophrenias may not be investigated for in most detained patients with a schizophrenia-like illness in England.
As secondary schizophrenias are present in 5%–8% of such cases, some of the variability in rates found by these authors must be related to the differing diagnostic rigour used to exclude secondary causes.
Citation: Hambidge D (2005) Secondary schizophrenia. PLoS Med 2(9): e279.
==== Refs
References
Saha S Chant D Welham J McGrath J A systematic review of the prevalence of schizophrenia PLoS Med 2005 2 e141 10.1371/journal.pmed.0020141 15916472
Hyde TM Lewis SW Hirsch SR Weinberger DR The secondary schizophrenias Schizophrenia 2003 Oxford Blackwell Publishing 832
Hambidge DM Detecting organic causes of first-episode psychosis Prog Neurol Psychiatry 2005 9 8 12
|
16173833
|
PMC1236793
|
CC BY
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e279
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020279
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383410.1371/journal.pmed.002028905-PLME-P-0294PerspectivesDiabetes/Endocrinology/MetabolismDrugs and Adverse Drug ReactionsEndocrinologyDiabetesInsulin Resistance in the Offspring of Parents with Type 2 Diabetes PerspectivesWagenmakers Anton J. M Anton J. M. Wagenmakers is Professor of Exercise Biochemistry, School of Sport and Exercise Sciences, University of Birmingham, Birmingham, United Kingdom. E-mail: [email protected]
Competing Interests: The author declares that no competing interests exist.
9 2005 27 9 2005 2 9 e289Copyright: © 2005 Anton J. M. Wagenmakers.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.
Decreased Insulin-Stimulated ATP Synthesis and Phosphate Transport in Muscle of Insulin-Resistant Offspring of Type 2 Diabetic Parents
Type 2 Diabetes: Insulin Resistance May Be the Result of Mitochondrial Dysfunction
Wagenmakers discusses the paper by Petersen and colleagues on insulin resistance in young lean individuals and its association with reduced phosphate transport into muscle cells and impaired mitochondrial energy generation in muscle.
==== Body
The global epidemic of type 2 diabetes is a pressing public health concern associated with a rapidly growing socioeconomic burden. Insulin resistance (IR) is an early event in the pathogenesis of type 2 diabetes. IR is characterised by a reduced ability of insulin to stimulate glucose uptake in skeletal muscle and by hyperglycaemia (high blood glucose concentration), at first only in periods following meal ingestion but later also in the overnight-fasted state.
In patients with type 2 diabetes, IR is progressive and after several years often leads to the development of secondary diabetes symptoms (including hypertriglyceridaemia, obesity, and pathology of the macro- and microvasculature). Eventually (after more than ten years) severe medical complications may develop, including retinopathy, neuropathy, tissue necrosis in the extremities, renal failure, and cardiomyopathy.
A progressive reduction over the last few decades of daily engagement in physically demanding activities and the recent growth in unbalanced diets are generally considered to be the primary causes of the dramatic rise in type 2 diabetes [1]. A genetic predisposition also runs in families and populations. In this month's PLoS Medicine, Kitt Petersen and colleagues [2] report new information on the early events in the underlying pathogenic mechanism that leads to the development of IR.
The Study's Findings
The authors investigated young, lean offspring with IR of parents with type 2 diabetes [2]. The reason to select offspring with IR is that a metabolic defect observed in this group is likely to be an early event of genetic origin and, therefore, is potentially a primary cause of the subsequent development of type 2 diabetes.
The offspring with IR were studied during a hyperinsulinaemic–euglycaemic clamp. This test is traditionally used to diagnose IR. The test measures the ability of insulin to stimulate the clearance of glucose from the blood during a simultaneous infusion of insulin in supraphysiological quantities and of glucose in quantities sufficient to maintain the glucose concentration at a normal physiological level. Petersen and colleagues found that infusion of insulin increased the turnover rate of adenosine triphosphate (ATP) in the skeletal muscle of control participants by 90%, while only a 5% increase (nonsignificant) was seen in offspring with IR. The increase in ATP turnover in the control participants means that the metabolic rate (energy expenditure) of the muscle goes up during the clamp study. In an earlier study by the same group of researchers [3], offspring with IR also were observed to have 30% lower rates of muscle ATP turnover in the overnight-fasted state.
What Do These Findings Mean?
ATP in resting fasted muscles is only produced for purposes of cell maintenance and survival functions (for example, maintenance of sodium and potassium gradients, amino acid gradients, protein synthesis rates, and functional organelles and membranes). Hence, Petersen et al.'s observations suggest that either the basal energy requirement is reduced in muscles of individuals with IR (potentially at the expense of the maintenance of cell functions) or the major control systems for mitochondrial respiration (simultaneous ATP synthesis and consumption) are not properly working (Figure 1).
Figure 1 Potential Mechanisms Leading to Failure of Insulin to Stimulate Muscle ATP Turnover during a Hyperinsulinaemic–Euglycaemic Clamp in Offspring with IR
The potential mechanisms are (1) a general mitochondrial dysfunction, reducing ATP production; (2) an impairment in the central nervous system (CNS) in the glucose- or insulin-induced excitation of muscle efferents, leading to reduced β-adrenergic activation of the muscle; (3) a reduced increase in mitochondrial ATP synthesis in response to activation of the β-adrenergic receptor (βAR); (4) a defect in insulin-induced opening of the terminal arterioles controlling blood flow through muscle fibre capillaries and, thus, preventing increases in the insulin concentration in the interstitial fluid and in binding of insulin to the insulin receptor (IR) in the muscle membrane; and (5) a molecular defect in the insulin signalling cascade in the muscle, leading to reduction in the insulin-induced stimulation of muscle glucose uptake, glycogen synthesis, and protein synthesis. ADP, adenosine diphosphate; NO, nitric oxide.
Petersen and colleagues favour the first explanation. They suggest that their combined observations point to a general mitochondrial dysfunction that impairs the ability of the mitochondria to synthesise ATP and oxidise fatty acids (FAs) at the normal resting rate, both in the basal overnight-fasted condition and after stimulation by insulin [2,3]. They also suggest that it is this mitochondrial dysfunction that causes IR.
The mitochondrial dysfunction is hypothesised to lead to a reduced ability to oxidise FAs and to the accumulation of triglycerides and FA metabolites (fatty AcylCoA, diacylglycerols, and ceramides). Such accumulation of triglycerides and FA metabolites has been repeatedly observed, both in skeletal muscle of obese individuals [4,5] with a reduced insulin sensitivity and in the muscle of healthy individuals given IR by the infusion of lipid emulsions and heparin [6]. These FA metabolites have been linked to the development of IR via a mechanism involving activation of protein kinase C and phosphorylation of the insulin receptor and IRS-1 at serine and threonine amino acid residues. Phosphorylation at these wrong amino acid residues prevents insulin-induced tyrosine phosphorylation of the insulin receptor and IRS-1, and prevents activation of the insulin signalling cascade; therefore, this mechanism is presently regarded as the primary cause of IR at the molecular level in the muscle [4–6].
Regular exercise and training should be considered interventions to correct the reduction in insulin-induced muscle ATP turnover.
But there is a problem with the assumption that a general mitochondrial dysfunction reduces both the basal and insulin-stimulated ATP production. In the muscle of both healthy control participants and patients with IR and type 2 diabetes, there is a large overcapacity in the ATP production capacity of skeletal muscle, allowing 5- to 20-fold increases in ATP turnover during exercise. Defects in resting mitochondrial ATP production, as occur in the muscle of patients with metabolic myopathies, lead to major reductions in the resting creatine phosphate/ATP ratio, and to a parallel increase in muscle lactate production as a consequence of a compensatory increase in the glycolytic ATP production. However, the offspring with IR in the previous studies by Petersen et al. [2,3] had normal resting creatine phosphate/ATP ratios and no change in muscle pH as a consequence of excessive lactate production. These findings seem to argue against the hypothesis that there is a general mitochondrial defect in offspring with IR and patients with type 2 diabetes.
Alternative Explanations for the Findings
An important open question concerns the mechanism by which insulin would increase the resting muscle ATP turnover rate in the control participants. Total energy expenditure in human tissues can be roughly broken down in three components: (1) obligatory energy expenditure required to perform cell maintenance and survival functions and maintain cell and body temperatures at 37 °C, (2) adaptive energy expenditure induced by nutrient ingestion, and (3) energy expenditure required to perform muscle contractions and physical activity [7–9]. Total energy expenditure is the sum of the energy required to perform all cellular and organ functions, plus heat production.
One possibility is that the increase in muscle ATP turnover, induced by insulin during the clamp, is caused by the well-known thermogenic effects of glucose and insulin [7–9]. Muscle glycogen synthesis and protein synthesis are energy-requiring metabolic processes, and both are stimulated when insulin binds to the insulin receptor in the muscle membrane. The stimulation of glycogen and protein synthesis leads to a need to increase muscle ATP turnover. As the efficiency of mitochondrial respiration is only 40%, increases in the rate of these metabolic reactions by definition contribute to heat production and the thermogenic effect of food ingestion. Therefore, insulin will increase muscle ATP production during the hyperinsulinaemic–euglycaemic clamp in healthy muscles in comparison to the basal state, while smaller increases in glycogen synthesis, in protein synthesis, and, therefore, in ATP production will occur in offspring with IR. This alternative mechanism explains the findings observed in Petersen et al. [2], but does not implicate that there is a pre-existent mitochondrial dysfunction.
The autonomic nervous system is known to modulate the thermogenic effect of glucose by activating small efferent nerves that end in the interstitium of skeletal muscle (the fluid that surrounds the muscle fibres). The nerve endings produce noradrenalin, which activates β-adrenoreceptors in the muscle membrane, and this leads to an increase in mitochondrial ATP production. The part of glucose-induced thermogenesis that is eliminated by β-adrenergic antagonists has been called “facultative thermogenesis” and is assumed to take place at least in part in skeletal muscle [9]. It has also been suggested [9] that insulin, via unidentified receptors, most probably located in the central nervous system, may stimulate muscle sympathetic nerve activity and facultative thermogenesis. The thermogenic effect of insulin and carbohydrates has been shown to be reduced in obese and insulin-resistant individuals [8,9]; therefore, an impairment in the mechanism leading to facultative thermogenesis may also explain a part of the reduction in insulin-stimulated muscle ATP synthesis observed in offspring with IR.
Recently, it has also been shown that insulin infusions lead to increases in blood flow through capillaries that surround the skeletal muscle fibres, both in healthy humans and rats [10–12]. The mechanism involves the binding of insulin to the insulin receptor on the endothelial cell layer that covers the luminar wall of the terminal arterioles that control blood flow through the muscle capillaries (Figure 1). This binding leads to activation of the insulin signalling cascade in the endothelial cells and to nitric oxide production [11–13]. Nitric oxide, a muscle relaxant, then diffuses to the smooth-muscle cell layer and leads to relaxation of the sphincter muscle, dilation of the terminal arterioles, and recruitment of muscle capillaries (Figure 1). The opening of the muscle capillaries in healthy control rats precedes the insulin-induced increase in glucose uptake, and the increase in glucose uptake can be prevented by prior infusion of nitric oxide synthase inhibitors. These observations suggest that insulin first recruits muscle capillaries before it can reach the insulin receptor in the muscle membrane and stimulate muscle glucose uptake, glycogen and protein synthesis, and ATP production. Severe defects do exist in this insulin-induced opening mechanism in obese insulin-resistant Zucker rats [14]. Although failure of the insulin-induced recruitment of muscle capillaries has not yet been shown to exist in humans with IR, it could also explain the reduced insulin-induced increase in muscle ATP synthesis rates in the offspring with IR observed by Petersen et al. [2], again without pointing to a mitochondrial defect or dysfunction.
The Study's Clinical Implications
Failure of insulin to stimulate muscle ATP production in offspring with IR may have multiple causes. A general mitochondrial dysfunction, as proposed by Petersen and colleagues, is one possibility, but the failure of insulin to (1) stimulate the insulin signalling cascade in muscle, (2) stimulate central thermogenic-control mechanisms of mitochondrial respiration, and (3) recruit muscle fibre capillaries are other potential mechanisms.
The basal observation that glucose and insulin do not stimulate muscle ATP production and thermogenesis in individuals with IR is clinically highly relevant, as it may explain the weight maintenance problems that people with IR experience. When there is a gradual reduction in the basal and insulin-induced energy expenditure at the muscle level during the development of type 2 diabetes, food intake should be reduced in proportion to the lower ATP need of the muscles. Failure to correct for the lower muscle energy requirement will lead to a positive energy balance and to weight gain. The data in Petersen and colleagues' study [2] also seem to suggest that the relative increase in energy expenditure by glucose and insulin is larger (90%) at the level of the muscle than at the level of the whole body (the whole-body thermogenic effect of orally ingested carbohydrates is maximally about 10%–15% [8,9]). A gradual disappearance of this large energy component in individuals with IR will lead to a substantially lower calorie and nutrient requirement.
Regular exercise and training should be considered interventions to correct the reduction in insulin-induced muscle ATP turnover. Endurance exercise performed three to four times per week may lead to more than 5-fold increases in the mitochondrial density (concentration) of a previously sedentary muscle [15], and will increase the ATP generating capacity. Both endurance and resistance exercise increase insulin sensitivity at the molecular level in the muscle, and they have also been suggested to increase the sensitivity of adrenergic control in both skeletal muscle and adipose tissue [15]. Exercise and training open muscle capillaries and increase glucose uptake in skeletal muscle by contraction-induced mechanisms that are independent of insulin action [12,14]. The measurement of muscle ATP turnover with magnetic resonance spectroscopy, as used in Petersen et al. [2], seems to be an ideal noninvasive method to investigate one critically important question: can changes towards a more active lifestyle reverse the observed reduction in insulin-induced muscle ATP turnover in the offspring with IR, and, in parallel, restore insulin sensitivity of muscle and precapillary arterioles and delay or prevent the later development of type 2 diabetes that was present in the parents?
Citation: Wagenmakers AJM (2005) Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med 2(9): e289.
Abbreviations
ATPadenosine triphosphate
FAfatty acid
IRinsulin resistance
==== Refs
References
World Health Organization Global strategy on diet, physical activity and health: Diabetes 2005 Geneva World Health Organization Available: http://www.who.int/dietphysicalactivity/publications/facts/diabetes/en/index.html . Accessed 27 July 2005
Petersen KF Dufour S Shulman GI Decreased insulin-stimulated rates of mitochondrial ATP synthesis and phosphate transport in skeletal muscle of insulin-resistant offspring of type 2 diabetic patients PLoS Med 2005 2 e233 10.1371/journal.pmed.0020233 16089501
Petersen KF Dufour S Befroy D Garcia R Shulman GI Impaired mitochondrial activity in the insulin-resistant offspring of patients with type 2 diabetes N Eng J Med 2004 350 664 671
Itani SI Ruderman NB Schmieder F Boden G Lipid-induced insulin resistance in human muscle is associated with changes in diacylglycerol, protein kinase C, and IkappaB-alpha Diabetes 2002 51 2005 2011 12086926
Hulver MW Dohm GL The molecular mechanisms linking muscle fat accumulation to insulin resistance Proc Nutr Soc 2004 63 375 380 15294058
Yu C Chen Y Cline GW Zhang D Zong H Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate-1 (IRS-1)-associated phosphatidylinositol 3-kinase activity in muscle J Biol Chem 2002 52 50230 50236
Lowell BB Bachman ES Beta-Adrenergic receptors, diet-induced thermogenesis, and obesity J Biol Chem 2003 278 29385 29388 12788929
De Jonge L Bray GA The thermic effect of food and obesity: A critical review Obesity Res 1997 5 622 631
Tappy L Thermic effect of food and sympathetic nervous system activity in humans Reprod Nutr Dev 1996 36 391 397 8878356
Coggins M Lindner J Rattigan S Jahn L Fasy E Physiologic hyperinsulinemia enhances human skeletal muscle perfusion by capillary recruitment Diabetes 2001 50 2682 2690 11723050
Vincent MA Clerk LH Lindner JR Klibanov AL Clark MG Microvascular recruitment is an early insulin effect that regulates skeletal muscle glucose uptake in vivo Diabetes 2004 53 1418 1423 15161743
Rattigan S Wheatly C Richards SM Barrett EJ Clark MG Exercise and insulin-mediated capillary recruitment in muscle Exerc Sport Sci Rev 2005 33 43 48 15640720
Kim F Tysseling KA Rice J Pham M Haji L Free fatty acid impairment of nitric oxide production in endothelial cells is mediated by IKKbeta Arterioscler Thromb Vasc Biol 2005 E-pub ahead of print
Wheatley CM Rattigan S Richards SM Barrett EJ Clark MG Skeletal muscle contraction stimulates capillary recruitment and glucose uptake in insulin-resistant obese Zucker rats Am J Physiol 2004 287 E804 E809
Zierath JR Hawley JA Skeletal muscle fibre type: Influence on contractile and metabolic properties PLoS Biol 2004 2 e348 10.1371/journal.pbio.0020348 15486583
|
16173834
|
PMC1236795
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e289
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020289
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383510.1371/journal.pmed.0020290PerspectivesBioinformatics/Computational BiologyOtherClinical PharmacologyPrimary CareGeneral MedicineDrugs and adverse drug reactionsChronic Disease ManagementGeneral Practice/Family Practice/Primary CareMedical errors/patient safetyQuality of health careComputerized Physician Order Entry Systems: The Coming of Age for Outpatient Medicine PerspectivesDavis Robert L Robert L. Davis is Senior Investigator, Center for Health Studies, Group Health Cooperative, Seattle, Washington, United States of America, and Professor of Epidemiology, University of Washington School of Public Health, Seattle, Washington, United States of America. E-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
9 2005 27 9 2005 2 9 e290Copyright: © 2005 Robert L. Davis.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.
Is it Possible to Change Prescribing Habits?
Davis discusses a new study in PLoS Medicine that evaluates the effectiveness of automated computerized alerts at reducing prescription errors in a primary care setting.
==== Body
In this month's issue of PLoS Medicine, Steele et al. address a timely question in clinical medicine today—how best to use the potential of computers to change and improve physicians' prescribing behavior [1]. In evaluating the effectiveness of automated computerized alerts in a primary-care setting, the authors examined a system that had been developed to reduce the chances of a physician prescribing a medication that would lead to any of the following five different drug-induced adverse events: hypokalemia, hyperkalemia, nephrotoxicity, thrombocytopenia, or hepatotoxicity. Specific rules triggered alerts that were delivered via computer to the prescribing physician whenever there existed a potential for one or more of these five adverse events. These alerts presented suggestions for the physician during the prescription-writing process; faced with an alert, a physician could elect to proceed, revise, or stop a medication order, or order additional laboratory tests.
The Study Design
Steele and colleagues used a “before and after” design to assess the impact of the intervention on physicians' prescribing behavior over a nine-month study period, during which 19,076 patients were seen. In the “before” phase, which lasted 17 weeks, the rules were switched on in the background without displaying any message to the providers. Since the rules were processing in the background, the providers did not receive any alerts recommending changes to their orders. Physicians' baseline ordering behavior was then compared to their ordering behavior in the “after” phase, during which alerts were presented to the provider. Adverse drug events were also assessed before and after the intervention by doing a random sample of chart reviews.
Can computers help to reduce prescribing errors?
Key Findings
During the six months after introduction of the intervention, almost 50% of all the medications prescribed triggered an assessment by the system. Alerts were presented to providers over 1,000 times (for about 6% of all the completed prescriptions). These alerts indicated a possible “lab value–prescription mismatch,” for instance, a lab test that was abnormal or absent, and the physician needed to account for this. After the intervention, there was a significant increase in the percentage of time that physicians ordered a rule-associated laboratory test (39% at baseline versus 51% post-intervention), particularly if the relevant laboratory value was not present in the data files. There was also a significant increase in the percentage of cases in which the provider stopped the ordering process and did not complete the medication order when an alert for an abnormal rule-associated laboratory result was displayed (5.6% versus 10.9%). There was also a nonsignificant decrease in the percentage of patients who had potential for an adverse drug event (10.3% versus 4.3%; p = 0.23).
Can Computers Help to Change Physicians' Behavior?
It has been quite difficult to work out how to develop interventions (using, for example, educational interventions or local experts and opinion leaders) that change physicians' behavior in ways that are generalizable, cost-efficient, clinically important, and persistent. Many of these same challenges exist with computers, where it has also been difficult to demonstrate important, clinically relevant, and lasting benefits from systems such as computerized alerts. The study by Steele et al. is important, as it demonstrates an ability to improve safety in prescribed medications, although it is equally important to acknowledge that this study showed little or no direct clinical benefit to patients.
This study adds further evidence to suggest that computerized systems for clinical medicine may be valuable. The study is a necessary step toward demonstrating the benefit of computerized physician order entry (CPOE) systems in the outpatient setting. To date, assessments of CPOE systems in the outpatient setting have been depressingly rare, with most studies of CPOE systems having been performed on patients in hospitals or in emergency departments. Since these clinical locations have had to deal with more complicated situations and more critically ill patients, it is not surprising that the early attempts by hospitals to automate their clinical information systems concentrated on such settings. There is now substantial evidence showing the benefit of different types of CPOE systems for hospitalized patients [2], but unfortunately there is little evidence for its effectiveness in the outpatient setting, where most prescriptions are written.
Studies of CPOE systems among critically ill hospitalized patients are not generalizable to the outpatient setting: not only do the types of prescriptions differ (typically oral for outpatients, while often intravenous or intramuscular for inpatients), but the patient issues differ as well. CPOE systems in the outpatient setting have to be geared mostly toward chronically ill patients with multiple needs centered on long-term management, whereas inpatient systems have to function for issues of acute illness in critically ill hospitalized patients. Hence, CPOE-based improvements in one setting cannot be assumed to extend to the other, since the conditions they deal with and the types of demands on the systems are so different.
Limitations of the Study
Although Steele and colleagues' study showed a favorable effect for the use of computerized alert systems, there were some limitations. Although there was an increase in laboratory tests ordered when previous test values were either abnormal or absent, the study was not powered—as the authors themselves note—to demonstrate a reduction in adverse drug events. Use of surrogate outcomes (such as changes in lab ordering behavior) is a well-accepted practice in studies where adverse outcomes are very rare and where such surrogates are linked to the occurrence of adverse outcomes. Nevertheless, before some clinics or practice groups decide to invest in expensive systems that will impact the day-to-day activities of many health-care providers, more studies of CPOE systems demonstrating not only a change in physician behavior but a real reduction in adverse events will likely be needed.
Also, because this study was not a randomized clinical trial, it is necessary to consider other reasons for the findings. In any “before and after” study such as this one, improvements in prescribing patterns over time may occur, regardless of whether or not the intervention was effective. External factors (such as statewide initiatives to improve prescription practices) or internal factors (such as increased attention by individual clinic physicians to the problem of drug–lab interactions) may lead to the erroneous observation that the intervention was effective when, in fact, improvements were a result of these other influences.
Assessments of CPOE systems in the outpatient setting have been depressingly rare.
A randomized clinical trial would have avoided the problems of trends over time and, if done correctly, could have reduced contamination by other factors. However, such trials are often lengthy and expensive. In comparison, many additions to CPOE systems are incremental, and interest in any particular change is often time-limited. Additionally, in many situations, CPOE systems are initiated by clinic-affiliated quality improvement departments, which consider numerous factors when deciding to purchase or develop such systems. In these cases, there may be little motivation to study individual aspects of such a system, much less undertake a costly randomized trial. Hence, for many changes, it will likely not be feasible to evaluate each one by way of randomized clinical trials.
Conclusion
Regardless of their limitations, observational studies such as the one by Steele et al. that compare nonrandomized physicians' behavior before interventions to behavior after interventions, and that depend largely on surrogate outcomes, provide powerful evidence on the effectiveness of computerized systems to reduce medication errors and improve patient safety. Similar studies will likely continue to be an important means used to evaluate many future developments in CPOE technology.
Citation: Davis RL (2005) Computerized physician order entry systems: The coming of age for outpatient medicine. PLoS Med 2(9): e290.
Abbreviation
CPOEcomputerized physician order entry
==== Refs
References
Steele AW Eisert S Witter J Lyons P Jones MA The effect of automated alerts on provider ordering behavior in an outpatient setting PLoS Med 2005 2 e255 10.1371/journal.pmed.0020255 16128621
Bates DW Cohen M Leape LL Overhage JM Shabot MM Reducing the frequency of errors in medicine using information technology J Am Med Inform Assoc 2001 8 299 308 11418536
|
16173835
|
PMC1236796
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e290
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020290
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0020300Correspondence and Other CommunicationsEpidemiology/Public HealthMental HealthPsychiatryEpidemiologyInternational healthPublic HealthSchizophrenia and other psychotic disordersAuthors' Reply: Measurement Errors in Schizophrenia Epidemiology CorrespondenceMcGrath John
1
Saha Sukanta
2
Welham Joy
2
Chant David Charles
2
1University of QueenslandWacol, QueenslandAustralia2Queensland Centre for Mental Health ResearchWacol, QueenslandAustraliaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
9 2005 27 9 2005 2 9 e300Copyright: © 2005 McGrath 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 Systematic Review of the Prevalence of Schizophrenia
Secondary Schizophrenia
==== Body
The letter from Hambidge highlights the heterogeneous nature of schizophrenia [1]. In order to diagnose schizophrenia, modern diagnostic criteria require the exclusion of other general somatic conditions that can mimic psychotic symptoms. Compliance with screening protocols designed to identify these disorders varies widely, even in developed countries. We agree with the correspondent that some studies included in our recent systematic review [2] would have probably included individuals who were subsequently found to have “secondary schizophrenia” (i.e., false positives). Thus, this issue would slightly inflate the prevalence estimate. The inappropriate inclusion of false positives is only one of a very long list of methodological factors that contribute to imprecision in the estimation of the incidence and prevalence of schizophrenia. The critical issue for the research community is how best to partition out measurement error from “true” variations in the incidence or prevalence of schizophrenia. In the absence of more refined phenotypes for the many different disorders that contribute to the syndrome of schizophrenia (e.g., by the use of yet-to-be-identified biomarkers), standard epidemiological studies of the incidence and prevalence of schizophrenia may have reached their limits of precision.
Citation: McGrath J, Saha S, Welham J, Chant DC (2005) Authors' reply: Measurement errors in schizophrenia epidemiology. PLoS Med 2(9): e300.
==== Refs
References
Hambidge D Secondary schizophrenia PLoS Med 2005 2 e279 10.1371/journal.pmed.0020279 16173833
Saha S Chant D Welham J McGrath J A systematic review of the prevalence of schizophrenia PLoS Med 2005 2 e141 10.1371/journal.pmed.0020141 15916472
|
0
|
PMC1236798
|
CC BY
|
2021-01-05 10:40:31
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e300
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020300
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383710.1371/journal.pmed.0020305PerspectivesObstetrics/GynecologyObstetricsDoes the Maxim “Once a Caesarean, Always a Caesarean” Still Hold True? PerspectivesUgwumadu Austin Austin Ugwumadu is a Consultant and Senior Lecturer in Obstetrics and Gynaecology at St. George's Hospital, London, United Kingdom. E-mail: [email protected]
Competing Interests: The author declares that he has no competing interests.
9 2005 27 9 2005 2 9 e305Copyright: © 2005 Austin Ugwumadu.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.
Predicting Cesarean Section and Uterine Rupture among Women Attempting Vaginal Birth after Prior Cesarean Section
Ugwumadu discusses a new tool, reported in PLoS Medicine, that can help to predict the risks of Caesarean section and uterine rupture in women attempting vaginal birth after prior Caesarean section.
A new tool can help to predict the risks of Caesarean section and uterine rupture in women attempting vaginal birth after prior Caesarean section
==== Body
Edwin Cragin's century-old opinion, “once a Caesarean, always a Caesarean,” is correct if placed in proper context [1]. Cragin was predicting the near certainty of repeat Caesarean section in a self-selected group of women (1%–2%) who failed to deliver vaginally after several days in active labour. At that time, rickets and pelvic deformity were prevalent even in industrialised countries, syntocinon for augmentation of slow labour was unknown, and surgery was crude and dangerous. The primary Caesarean section was undertaken to save the life of an exhausted, dehydrated, ketotic, often pyrexial and delirious, moribund mother. In those days, fetal compromise was not an indication for Caesarean section; indeed there was no such thing as fetal monitoring (either antepartum or intrapartum). Cragin recognised that women who survived one Caesarean section were not candidates for vaginal delivery in subsequent pregnancies.
The Rising Rate of Caesarean Sections
Within the last three decades, Caesarean section rates in many countries have risen 5-fold to 10-fold. For example, in England and Wales the rate rose from 4% in 1970 to a current 22% [2]. In the United States it rose from 6% in 1970 to 24% in 1990 [3], while in the municipality of Ribeirao Preto, State of Sao Paulo, Brazil, the rate rose from 30.3% in 1978–1979 to 50.8% in 1994 [4]. The rising rate has been driven at least in part by our reliance on electronic fetal monitoring, pressure from health consumers to salvage small babies even at the very margins of viability, fear of litigation, decreasing expertise in operative vaginal deliveries and, in the West, lifestyle choices.
The current Caesarean section rate of about 50% in some countries [4] is too high and unsustainable, and according to the World Health Organization, is not associated with any further improvement in perinatal outcome compared to outcomes at a Caesarean section rate of 10%–15% [5]. Can we halt and reverse this trend, reduce the morbidity and drain on health care budgets associated with it, and, above all, balance maternal choice issues? In contrast to Edwin Cragin's patients, women today are healthier (rickets is rare now, even in nonindustrialised countries), and in the rich world at least, oxytocin, blood transfusion, antibiotics, and thromboprophylaxis are available, while surgery and anaesthesia are safe. Therefore, some obstetricians have enthusiastically and nonselectively promoted vaginal birth after Caesarean section. However, the consequences in inappropriate cases can be disastrous.
The Risks of Labour after Caesarean Section
Labour/vaginal birth after Caesarean section is associated with increased risks of uterine rupture and feto-maternal morbidity and mortality. These risks and costs of care rise further if the attempt fails [6]. Available evidence suggests that the complication rates are lowest in women whose attempt at vaginal delivery after Caesarean has been successful—even lower than in women who had planned Caesarean section. Therefore, the crucial questions are how to reliably predict successful attempt at vaginal birth after Caesarean section, and how to determine and quantify the magnitude of the risk of failure that is acceptable to women and their caregivers.
In this issue of PLoS Medicine, Smith and colleagues aimed to develop a prediction tool to predict the likelihood of Caesarean section and uterine rupture in women with one prior Caesarean undergoing a trial of labour [7]. Such a tool would be useful for counseling women and for policy makers and health care commissioners. The authors randomly allocated 23,286 women (from the linked Scottish Morbidity Record [SMR2] and the Scottish Stillbirth and Neonatal Death Enquiry) to two groups: a model development group and a model validation group. None of the previously published prediction tools had been validated prospectively. The authors' model ranked women as high risk of emergency Caesarean section (≥ 40%), or low risk (≤ 20%), and the primary analysis was confined to women who delivered ≥ 40 weeks' gestation.
The factors that were associated with Caesarean section, based on the multivariate analysis, are shown in Table 1. The predicted risk of Caesarean section was also associated with the risk of uterine rupture (odds ratio [OR] 1.22, for a 5% increase in predicted risk, 95% confidence interval [CI] 1.14–1.31) and rupture associated with perinatal death (OR 1.32 for a 5% increase in predicted risk, 95% CI 1.02–1.73). In the validation group, 10.9% of the women predicted to have a low risk of having Caesarean section actually had a Caesarean section, while 47.7% of women predicted to have a high risk of Caesarean section had one. The incidence of uterine rupture was 2.0 and 9.1 per 1000 in the low and high-risk categories respectively (OR 4.5; 95% CI 2.6–8.1).
Table 1 Factors Associated with Caesarean Section in Women with One Prior Caesarean Undergoing a Trial of Labour
Data taken from the multivariate analysis in [7].
The Value of the New Predictive Tool
When applied to the total population, the predictive model classified just over half (52.5%) of the study population into a low (36%) and a high (16.5%) likelihood of Caesarean. Over half of those predicted to be at high risk of Caesarean section did not have one, suggesting that the value of this tool may well lie in predicting those patients who are unlikely to have an emergency Caesarean and/or uterine rupture (with or without perinatal death).
At best, the tool classifies about half our obstetric population.
By confining the primary analysis to women who delivered at ≥ 40 weeks' gestation, the authors aimed to capture only the women who “truly” intended to have a vaginal delivery, assuming that planned Caesarean deliveries would have been undertaken by 39 weeks' gestation. There is some evidence, however, that the likelihood of successful vaginal delivery decreases beyond 40 weeks' gestation in women with prior Caesarean section [8,9], raising the question of how generalisable these findings are to a significant proportion of women with previous Caesarean section who go into spontaneous labour between 37 and 40 weeks' gestation. Not surprisingly, when the model was applied to this latter group in a secondary analysis, the risk of Caesarean section was found to be 18%, a rate predictably lower than the chosen cutoff for low risk of Caesarean section. It would seem reasonable, therefore, to discuss this low risk with women and indeed to encourage them to proceed with a trial of vaginal delivery if spontaneous labour supervenes before 40 weeks.
Even in the high-risk group, the risk of uterine rupture was less than 1% (9.1 per 1000). One question is whether this magnitude of risk justifies a repeat Caesarean section in modern units that deploy round-the-clock obstetric, pathology, and anaesthetic services, practise continuous electronic fetal monitoring, and have facilities for emergency Caesarean section.
The Issue of Women's Choice
The more fundamental 21st century question is whether a previous Caesarean section is a medical indication for a repeat Caesarean. If the answer is “yes,” then women can truly exercise the choice to have or not to have one. If the answer is “no,” then true choice hardly exists, at least not in the public sector. The right to choose a Caesarean may then adversely affect the rights and just expectations of other women and their babies who medically need operative delivery. Autonomous clinicians, conscious of responsible utilisation of health care resources, will decline what is arguably then a lifestyle choice. Guidelines from The National Institute for Clinical Excellence (http://www.nice.org.uk) on Caesarean section recommend that “maternal request is not on its own an indication for Caesarean section,” and that “a clinician can decline such a request” [10]. The guidelines, however, remain silent on how this applies to women with previous Caesarean section. For example, are obstetricians liable if they persuade a woman to have a trial of labour, and complications occur?
Smith et al. should be applauded for developing their new tool for the prediction of Caesarean section and uterine rupture in women with previous Caesarean. It is a huge step forward, but it is not the definitive tool. At best, it classifies just about half our obstetric population. The critical questions of women's choice and the medical view of labour after Caesarean section in the 21st century remain unanswered.
Citation: Ugwumadu A (2005) Does the maxim “Once a Caesarean, always a Caesarean” still hold true? PLoS Med 2(9): e305.
Abbreviations
CIconfidence interval
ORodds ratio
==== Refs
References
Cragin EB Conservatism in obstetrics NY Med J 1916 104 1 3
RCOG Clinical Effectiveness Support Unit The National Sentinel caesarean section audit report 2001 Available: http://www.rcog.org.uk/resources/public/pdf/nscs_audit.pdf . Accessed 1 August 2005
Notzon FC Cnattingius S Bergsjo P Cole S Taffel S Cesarean section delivery in the 1980s: International comparison by indication Am J Obstet Gynecol 1994 170 495 504 8116703
Gomes UA Silva AA Bettiol H Barbieri MA Risk factors for the increasing caesarean section rate in Southeast Brazil: A comparison of two birth cohorts, 1978–1979 and 1994 Int J Epidemiol 1999 28 687 694 10480697
World Health Organization Appropriate technology for birth Lancet 1985 2 436 437 2863457
Rosen MG Dickinson JC Westhoff CL Vaginal birth after cesarean: A meta-analysis of morbidity and mortality Obstet Gynecol 1991 77 465 470 1825136
Smith GCS White IR Pell JP Dobbie R Predicting cesarean section and uterine rupture among women attempting vaginal birth after prior cesarean section PLoS Med 2005 2 e252 10.1371/journal.pmed.0020252 16146414
Yeh S Huang X Phelan JP Postterm pregnancy after previous cesarean section J Reprod Med 1984 29 41 44 6708020
Zelop CM Shipp TD Cohen A Repke JT Lieberman E Trial of labor after 40 weeks' gestation in women with prior cesarean Obstet Gynecol 2001 97 391 393 11239643
National Institute for Clinical Excellence Caesarean section Clinical guideline 13 2004 Available: http://www.nice.org.uk/pdf/CG013NICEguideline.pdf . Accessed 1 August 2005
|
16173837
|
PMC1236800
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e305
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020305
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383810.1371/journal.pmed.0020307Correspondence and Other CommunicationsInfectious DiseasesPrimary CareGeneral Practice/Family Practice/Primary CareInfectious DiseasesMisleading Information on the Properties of Vitamin C CorrespondenceHickey Steve
1
Roberts Hilary
2
1Manchester Metropolitan UniversityManchesterUnited Kingdom2ManchesterUnited KingdomE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
9 2005 27 9 2005 2 9 e307Copyright: © 2005 Hickey and Roberts.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.
Vitamin C For Preventing and Treating the Common Cold
Authors' Reply
Narrow Scope of Vitamin C Review
==== Body
The Cochrane review by Douglas et al. [1], which is referenced in the Best Practice article by Douglas and Hemilä [2], covers 60 years of research into vitamin C and the common cold. However, the review omits pharmacokinetic data that invalidate the conclusion that vitamin C is ineffective. This conclusion is not derivable from the data presented.
The dual-phase pharmacokinetics of vitamin C are described by the dynamic flow model [3,4]. Low gram-level intakes of ascorbate, leading to blood plasma levels below 70 μM/l, have a half-life of 8–40 days. Higher gram-level intakes have a plasma half-life of 30 minutes [3]. A large oral dose raises blood plasma levels briefly: they reach a peak after two to three hours, before decaying back to baseline. Frequent repeated doses allow sustained high plasma levels of about 250 μM/l [4,5].
Douglas and Hemilä reviewed intakes that transiently raise plasma ascorbate levels above 70 μM/l. A single dose does not raise the median level [6,7]. Daily supplements would, thus, not increase disease resistance to any great degree [3,4]. Single or double doses daily will not increase background plasma levels, regardless of the magnitude of the dose [6,7]. Since plasma ascorbate is at background level for the majority of the day, effects will be minimal.
There is widespread confusion about nutritional and pharmacological levels of supplementation [3]. Linus Pauling, typically, described nutritional gram-level doses able to provide a degree of disease prevention [8]. By contrast, pharmacological doses used for treatment are, at minimum, an order of magnitude larger and involve frequent doses. The doses should be at intervals of three hours or less [3]. Treatment doses are described by Cathcart's paper on titration to bowel tolerance [9]. To treat the onset of a cold, the therapy is perhaps a minimum of 10 g of oral ascorbic acid, followed by at least 2 g each hour [3,4].
Douglas and Hemilä give a misleading impression by not making it clear that the doses they consider are not pharmacological. They claim that the results of one study, giving an 8-g dose at the start of symptoms, are tantalising and deserve further assessment. However, once this single dose has been excreted, the protective effects will be lost. During illness, ascorbate is depleted rapidly and higher oral intakes are tolerated—up to 200 g per day [9]. It would be surprising if this 8-g dose had a large effect.
Studies on ascorbate require appropriate doses. Douglas and Hemilä have only confirmed that 60 years of vitamin C research has largely been wasted because of confusion between nutritional and pharmacological intakes, and because of a misunderstanding of the pharmacokinetics. It is essential that high-dose studies take into account ascorbate's dual-phase pharmacokinetics. The dosing regime should allow sustained high plasma levels to be achieved. The claim that vitamin C cannot prevent or cure the common cold is both premature and unwarranted.
Citation: Hickey S, Roberts H (2005) Misleading information on the properties of vitamin C. PLoS Med 2(9): e307.
==== Refs
References
Douglas RM Hemilä H D'Souza R Chalker EB Treacy B Vitamin C for preventing and treating the common cold Cochrane Database Syst Rev 2004 4 CD000980 pub2
Douglas RM Hemilä H Vitamin C for preventing and treating the common cold PLoS Med 2005 2 e168 10.1371/journal.pmed.0020168 15971944
Hickey S Roberts HJ Ascorbate: The science of vitamin C 2004 Napa (California) Lulu Press 264
Hickey S Roberts HJ Cathcart RF Dynamic flow: A new model for ascorbate J Orthomol Med 2005 In press
Padayatty SJ Sun H Wang Y Riordan HD Hewitt SM Vitamin C pharmacokinetics: Implications for oral and intravenous use Ann Intern Med 2004 140 533 537 15068981
Levine M Conry-Cantilena C Wang Y Welch RW Washko PW Vitamin C pharmacokinetics in healthy volunteers: Evidence for a recommended dietary allowance Proc Natl Acad Sci U S A 1996 93 3704 3709 8623000
Levine M Wang Y Padayatty SJ Morrow J A new recommended dietary allowance of vitamin C for healthy young women Proc Natl Acad Sci U S A 2001 98 9842 9846 11504949
Pauling L Vitamin C and the common cold 1970 New York W. H. Freeman 122
Cathcart RF Vitamin C, titrating to bowel tolerance, anascoremia, and acute induced scurvy Med Hypotheses 1981 7 1359 1376 7321921
|
16173838
|
PMC1236801
|
CC BY
|
2021-01-05 10:40:31
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e307
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020307
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617383810.1371/journal.pmed.0020307Correspondence and Other CommunicationsInfectious DiseasesPrimary CareGeneral Practice/Family Practice/Primary CareInfectious DiseasesMisleading Information on the Properties of Vitamin C CorrespondenceHickey Steve
1
Roberts Hilary
2
1Manchester Metropolitan UniversityManchesterUnited Kingdom2ManchesterUnited KingdomE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
9 2005 27 9 2005 2 9 e307Copyright: © 2005 Hickey and Roberts.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.
Vitamin C For Preventing and Treating the Common Cold
Authors' Reply
Narrow Scope of Vitamin C Review
==== Body
The Cochrane review by Douglas et al. [1], which is referenced in the Best Practice article by Douglas and Hemilä [2], covers 60 years of research into vitamin C and the common cold. However, the review omits pharmacokinetic data that invalidate the conclusion that vitamin C is ineffective. This conclusion is not derivable from the data presented.
The dual-phase pharmacokinetics of vitamin C are described by the dynamic flow model [3,4]. Low gram-level intakes of ascorbate, leading to blood plasma levels below 70 μM/l, have a half-life of 8–40 days. Higher gram-level intakes have a plasma half-life of 30 minutes [3]. A large oral dose raises blood plasma levels briefly: they reach a peak after two to three hours, before decaying back to baseline. Frequent repeated doses allow sustained high plasma levels of about 250 μM/l [4,5].
Douglas and Hemilä reviewed intakes that transiently raise plasma ascorbate levels above 70 μM/l. A single dose does not raise the median level [6,7]. Daily supplements would, thus, not increase disease resistance to any great degree [3,4]. Single or double doses daily will not increase background plasma levels, regardless of the magnitude of the dose [6,7]. Since plasma ascorbate is at background level for the majority of the day, effects will be minimal.
There is widespread confusion about nutritional and pharmacological levels of supplementation [3]. Linus Pauling, typically, described nutritional gram-level doses able to provide a degree of disease prevention [8]. By contrast, pharmacological doses used for treatment are, at minimum, an order of magnitude larger and involve frequent doses. The doses should be at intervals of three hours or less [3]. Treatment doses are described by Cathcart's paper on titration to bowel tolerance [9]. To treat the onset of a cold, the therapy is perhaps a minimum of 10 g of oral ascorbic acid, followed by at least 2 g each hour [3,4].
Douglas and Hemilä give a misleading impression by not making it clear that the doses they consider are not pharmacological. They claim that the results of one study, giving an 8-g dose at the start of symptoms, are tantalising and deserve further assessment. However, once this single dose has been excreted, the protective effects will be lost. During illness, ascorbate is depleted rapidly and higher oral intakes are tolerated—up to 200 g per day [9]. It would be surprising if this 8-g dose had a large effect.
Studies on ascorbate require appropriate doses. Douglas and Hemilä have only confirmed that 60 years of vitamin C research has largely been wasted because of confusion between nutritional and pharmacological intakes, and because of a misunderstanding of the pharmacokinetics. It is essential that high-dose studies take into account ascorbate's dual-phase pharmacokinetics. The dosing regime should allow sustained high plasma levels to be achieved. The claim that vitamin C cannot prevent or cure the common cold is both premature and unwarranted.
Citation: Hickey S, Roberts H (2005) Misleading information on the properties of vitamin C. PLoS Med 2(9): e307.
==== Refs
References
Douglas RM Hemilä H D'Souza R Chalker EB Treacy B Vitamin C for preventing and treating the common cold Cochrane Database Syst Rev 2004 4 CD000980 pub2
Douglas RM Hemilä H Vitamin C for preventing and treating the common cold PLoS Med 2005 2 e168 10.1371/journal.pmed.0020168 15971944
Hickey S Roberts HJ Ascorbate: The science of vitamin C 2004 Napa (California) Lulu Press 264
Hickey S Roberts HJ Cathcart RF Dynamic flow: A new model for ascorbate J Orthomol Med 2005 In press
Padayatty SJ Sun H Wang Y Riordan HD Hewitt SM Vitamin C pharmacokinetics: Implications for oral and intravenous use Ann Intern Med 2004 140 533 537 15068981
Levine M Conry-Cantilena C Wang Y Welch RW Washko PW Vitamin C pharmacokinetics in healthy volunteers: Evidence for a recommended dietary allowance Proc Natl Acad Sci U S A 1996 93 3704 3709 8623000
Levine M Wang Y Padayatty SJ Morrow J A new recommended dietary allowance of vitamin C for healthy young women Proc Natl Acad Sci U S A 2001 98 9842 9846 11504949
Pauling L Vitamin C and the common cold 1970 New York W. H. Freeman 122
Cathcart RF Vitamin C, titrating to bowel tolerance, anascoremia, and acute induced scurvy Med Hypotheses 1981 7 1359 1376 7321921
|
16173839
|
PMC1236802
|
CC BY
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e308
|
latin-1
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020308
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0020309Correspondence and Other CommunicationsInfectious DiseasesPrimary CareGeneral Practice/Family Practice/Primary CareInfectious DiseasesAuthors' Reply CorrespondenceHemilä Harri
1
Douglas Robert M
2
1University of HelsinkiHelsinkiFinland2Australian National UniversityCanberraAustraliaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
9 2005 27 9 2005 2 9 e309Copyright: © 2005 Hemilä and Douglas.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.
Vitamin C For Preventing and Treating the Common Cold
Narrow Scope of Vitamin C Review
Misleading Information on the Properties of Vitamin C
==== Body
The responses to our Best Practice article [1] by Hickey and Roberts [2], and by Sardi [3], make the same point, namely, that a recent pharmacokinetic study reported that frequent oral intakes of vitamin C would be necessary to elevate plasma ascorbic acid levels to the point where they believe it would have a pharmacological impact. Both authors suggest that the conclusions of our Cochrane review [4] are flawed because all of the placebo-controlled trials that have been carried out so far have used, for both prophylaxis and therapy, one to three doses per day of vitamin C, ranging from 200 mg daily to as much as 8 g in a single daily dose.
We have not, as our critics imply, concluded that vitamin C, in the doses used in trials reported in the literature, has no effect on the common cold. On the contrary, our evidence indicated that in marathon runners and in those exposed to high physical or cold stress, a substantial prophylactic effect was observed. And in the general population using regular vitamin C prophylaxis, common cold duration was consistently shortened, but the level of shortening was relatively trivial.
We do not consider the vitamin C and the common cold relationship closed. Nor are we persuaded by the arguments of these three critics that frequent, large doses would necessarily result in substantially greater benefits than earlier trials have demonstrated.
We consider that it may be useful to distinguish between (a) prophylactic supplementation for people who are in good health and (b) therapeutic supplementation for people who have an infection. The kidneys reabsorb essentially all vitamin C when the dietary intake is below 60–100 mg/day, and the vitamin C level in leukocytes is saturated by approximately 100 mg/day [5]; in this respect, we doubt that prophylactic supplementation of healthy people, using doses higher than those in the published trials, might be expected to benefit the general healthy population. On the other hand, there is evidence indicating that common cold infection decreases the vitamin C level in leukocytes, suggesting changes in vitamin C metabolism [6], and, in this respect, there seems to be a rationale to study the effects of supplementation on people infected with the common cold using even higher doses.
To this point, the claim that these two letters make has not been reported in properly conducted randomized controlled trials of either therapy or prophylaxis. We look forward to incorporating such trials, when they have been carried out, in future versions of the Cochrane review. Meanwhile, we stand firmly by the conclusions reported in our article.
Citation: Hemilä H, Douglas RM (2005) Authors' reply. PLoS Med 2(9): e309.
==== Refs
References
Douglas RM Hemilä H Vitamin C for preventing and treating the common cold PLoS Med 2005 2 e168 10.1371/journal.pmed.0020168 15971944
Hickey S Roberts H Misleading information on the properties of vitamin C PLoS Med 2005 2 e307 10.1371/journal.pmed.0020307 16173838
Sardi W Narrow scope of vitamin C review PLoS Med 2005 2 e308 10.1371/journal.pmed.0020308 16173839
Douglas RM Hemilä H D'Souza R Chalker EB Treacy B Vitamin C for preventing and treating the common cold Cochrane Database Syst Rev 2004 4 CD000980 pub2
Levine M Conry-Cantilena C Wang Y Welch RW Washko PW Vitamin C pharmacokinetics in healthy volunteers: Evidence for a recommended dietary allowance Proc Natl Acad Sci U S A 1996 93 3704 3709 8623000
Hume R Weyers E Changes in leucocyte ascorbic acid during the common cold Scott Med J 1973 18 3 7 4717661
|
0
|
PMC1236803
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e309
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020309
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617384110.1371/journal.pmed.0020310Correspondence and Other CommunicationsOtherPrimary CareGeneral MedicineCommunication in Health CareGeneral Practice/Family Practice/Primary CarePatientsTamoxifen and the Singing Voice CorrespondenceHerxheimer Andrew
1
1United Kingdom Cochrane CentreOxfordUnited KingdomE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
9 2005 27 9 2005 2 9 e310Copyright: © 2005 Andrew Herxheimer.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.
Communicating with Patients about Harms and Risks
==== Body
My remark, in my recent Essay in PLoS Medicine [1], that deepening of the voice occurs with long-term use of tamoxifen for breast cancer needs qualification.
Several colleagues have rightly pointed out that the evidence for the effect is less clear than I implied: it comes from women who have experienced it [2], but there have been no controlled studies. A change in voice was looked for and not found among effects spontaneously reported in large trials of tamoxifen, but this was not specifically asked about and might well have been missed. It is also recognised that the voice sometimes becomes deeper during or after menopause, in the absence of tamoxifen.
To convey the uncertainty of the facts, I wish to amend my statement as follows: “The irreversible deepening of the voice that has been reported to occur with long-term use of tamoxifen for breast cancer is an example of a side effect that prescribers, manufacturers, and drug regulators seem to have considered trivial and have not investigated.”
Citation: Herxheimer A (2005) Tamoxifen and the singing voice. PLoS Med 2(9): e310.
==== Refs
References
Herxheimer A Communicating with patients about harms and risks PLoS Med 2005 2 e42 10.1371/journal.pmed.0020042 15736998
Goodare H Tamoxifen and singing Breast Care and Mastectomy (BCMA) Network 1992 1992 2 (Spring)
|
16173841
|
PMC1236804
|
CC BY
|
2021-01-05 10:40:33
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e310
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020310
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617384210.1371/journal.pmed.0020311Correspondence and Other CommunicationsDiabetes/Endocrinology/MetabolismEpidemiology/Public HealthGeneral MedicineChronic Disease ManagementEpidemiologyObesityResponse to Stampfer Commentary CorrespondenceWilliamson David F
1
1Centers for Disease Control and PreventionAtlanta, GeorgiaUnited States of AmericaE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
9 2005 27 9 2005 2 9 e311Copyright: © 2005 David F. Williamson.2005This 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.
Weight Loss and Mortality: What Does the Evidence Show?
Author's Reply
==== Body
Stampfer's recent Perspective [1] on the paper by Sørensen et al. [2] appropriately acknowledges the challenges inherent in using observational epidemiology to determine the impact of weight loss on life expectancy. However, his case that the data of Sørensen et al. do not support their conclusion that intentional weight loss may be hazardous is based, in part, on erroneous statements about the study.
Stampfer suggests that “reverse causation” could account for the findings of Sørensen et al. because he believes they did not do a “lagged” analysis in which deaths that occur in the first few years after follow-up are excluded. In the statistical analysis, however, Sørensen et al. describe using two separate fully adjusted models: one for the first five years of follow-up and one for the period thereafter, and they also reported mortality hazard ratios (HRs) associated with intentional weight loss during each period. Because so few deaths occurred in the first five years of follow-up, the estimated mortality HR for intentional weight loss during this period (6.26) had such a wide confidence interval (0.33–118) that it was essentially meaningless. However, after excluding the first five years of follow-up data, Sørensen et al. still found a clinically and statistically significant association between intentional weight loss and death during the remaining 13 years of follow-up: HR = 1.88 (confidence interval, 1.05–3.39).
Stampfer indicates that the authors differentiated only between current smokers and nonsmokers and, thus, inappropriately combined never smokers with past smokers. In their methods, however, Sørensen et al. reported that they originally used four categories (never smoker, occasional smoker, former regular smoker, and current smoker) to code the smoking status of the study's participants, before recoding smoking status as a dichotomous yes-or-no variable. However, as Sørensen et al. described in their statistical analysis, they analyzed their models using both of the coding methods to determine whether recoding resulted in residual confounding. Because they found no residual confounding, they chose to report results only from the model with the simpler, dichotomous coding of smoking status.
Stampfer also argues that the best way to remove residual confounding by smoking is to “simply exclude current and past smokers” [1]. This exclusionary approach for smoking has been previously examined in a methodological study that utilized statistical simulation, with data from 15 diverse observational studies of body weight and mortality [3]. The study concluded that eliminating smokers from the datasets prior to analysis produces results similar to those expected from the elimination of numerically similar random proportions of the datasets prior to analysis [3]. Thus, the practice of excluding smokers in studies of weight loss and mortality is highly questionable.
Citation: Williamson DF (2005) Response to Stampfer commentary. PLoS Med 2(9): e311.
==== Refs
References
Stampfer M Weight loss and mortality: What does the evidence show? PLoS Med 2005 2 e181 10.1371/journal.pmed.0020181 15971953
Sørensen TIA Rissanen A Korkeila M Kaprio J Intention to lose weight, weight changes, and 18-y mortality in overweight individuals without comorbidities PLoS Med 2005 2 e171 10.1371/journal.pmed.0020171 15971946
BMI in Diverse Populations Collaborative Group Effect of smoking on the body mass index–mortality relation: Empirical evidence from 15 studies Am J Epidemiol 1999 150 1297 1308 10604772
|
16173842
|
PMC1236805
|
CC0
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e311
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020311
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1617384210.1371/journal.pmed.0020311Correspondence and Other CommunicationsDiabetes/Endocrinology/MetabolismEpidemiology/Public HealthGeneral MedicineChronic Disease ManagementEpidemiologyObesityResponse to Stampfer Commentary CorrespondenceWilliamson David F
1
1Centers for Disease Control and PreventionAtlanta, GeorgiaUnited States of AmericaE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
9 2005 27 9 2005 2 9 e311Copyright: © 2005 David F. Williamson.2005This 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.
Weight Loss and Mortality: What Does the Evidence Show?
Author's Reply
==== Body
Stampfer's recent Perspective [1] on the paper by Sørensen et al. [2] appropriately acknowledges the challenges inherent in using observational epidemiology to determine the impact of weight loss on life expectancy. However, his case that the data of Sørensen et al. do not support their conclusion that intentional weight loss may be hazardous is based, in part, on erroneous statements about the study.
Stampfer suggests that “reverse causation” could account for the findings of Sørensen et al. because he believes they did not do a “lagged” analysis in which deaths that occur in the first few years after follow-up are excluded. In the statistical analysis, however, Sørensen et al. describe using two separate fully adjusted models: one for the first five years of follow-up and one for the period thereafter, and they also reported mortality hazard ratios (HRs) associated with intentional weight loss during each period. Because so few deaths occurred in the first five years of follow-up, the estimated mortality HR for intentional weight loss during this period (6.26) had such a wide confidence interval (0.33–118) that it was essentially meaningless. However, after excluding the first five years of follow-up data, Sørensen et al. still found a clinically and statistically significant association between intentional weight loss and death during the remaining 13 years of follow-up: HR = 1.88 (confidence interval, 1.05–3.39).
Stampfer indicates that the authors differentiated only between current smokers and nonsmokers and, thus, inappropriately combined never smokers with past smokers. In their methods, however, Sørensen et al. reported that they originally used four categories (never smoker, occasional smoker, former regular smoker, and current smoker) to code the smoking status of the study's participants, before recoding smoking status as a dichotomous yes-or-no variable. However, as Sørensen et al. described in their statistical analysis, they analyzed their models using both of the coding methods to determine whether recoding resulted in residual confounding. Because they found no residual confounding, they chose to report results only from the model with the simpler, dichotomous coding of smoking status.
Stampfer also argues that the best way to remove residual confounding by smoking is to “simply exclude current and past smokers” [1]. This exclusionary approach for smoking has been previously examined in a methodological study that utilized statistical simulation, with data from 15 diverse observational studies of body weight and mortality [3]. The study concluded that eliminating smokers from the datasets prior to analysis produces results similar to those expected from the elimination of numerically similar random proportions of the datasets prior to analysis [3]. Thus, the practice of excluding smokers in studies of weight loss and mortality is highly questionable.
Citation: Williamson DF (2005) Response to Stampfer commentary. PLoS Med 2(9): e311.
==== Refs
References
Stampfer M Weight loss and mortality: What does the evidence show? PLoS Med 2005 2 e181 10.1371/journal.pmed.0020181 15971953
Sørensen TIA Rissanen A Korkeila M Kaprio J Intention to lose weight, weight changes, and 18-y mortality in overweight individuals without comorbidities PLoS Med 2005 2 e171 10.1371/journal.pmed.0020171 15971946
BMI in Diverse Populations Collaborative Group Effect of smoking on the body mass index–mortality relation: Empirical evidence from 15 studies Am J Epidemiol 1999 150 1297 1308 10604772
|
0
|
PMC1236811
|
CC BY
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2005 Sep 27; 2(9):e326
|
latin-1
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020326
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1619309510.1371/journal.pcbi.001004005-PLCB-RA-0073R4plcb-01-04-05Research ArticleBioinformatics - Computational BiologyGenetics/Functional GenomicsPlant ScienceEukaryotesPlantsArabidopsisInferring Hypotheses on Functional Relationships of Genes: Analysis of the Arabidopsis thaliana Subtilase Gene Family Arabidopsis Subtilases
Rautengarten Carsten 1*Steinhauser Dirk 2Büssis Dirk 1Stintzi Annick 3Schaller Andreas 3Kopka Joachim 2Altmann Thomas 11 Institut für Biochemie und Biologie, Genetik, Universität Potsdam, Golm, Germany
2 Max-Planck-Institut für Molekulare Pflanzenphysiologie, Golm, Germany
3 Institut für Physiologie und Biotechnologie der Pflanzen, Universität Hohenheim, Stuttgart, Germany
Alexandrov Nikolai N EditorCeres Inc., United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 23 9 2005 1 4 e4013 4 2005 16 8 2005 Copyright: © 2005 Rautengarten 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 gene family of subtilisin-like serine proteases (subtilases) in Arabidopsis thaliana comprises 56 members, divided into six distinct subfamilies. Whereas the members of five subfamilies are similar to pyrolysins, two genes share stronger similarity to animal kexins. Mutant screens confirmed 144 T-DNA insertion lines with knockouts for 55 out of the 56 subtilases. Apart from SDD1, none of the confirmed homozygous mutants revealed any obvious visible phenotypic alteration during growth under standard conditions. Apart from this specific case, forward genetics gave us no hints about the function of the individual 54 non-characterized subtilase genes. Therefore, the main objective of our work was to overcome the shortcomings of the forward genetic approach and to infer alternative experimental approaches by using an integrative bioinformatics and biological approach. Computational analyses based on transcriptional co-expression and co-response pattern revealed at least two expression networks, suggesting that functional redundancy may exist among subtilases with limited similarity. Furthermore, two hubs were identified, which may be involved in signalling or may represent higher-order regulatory factors involved in responses to environmental cues. A particular enrichment of co-regulated genes with metabolic functions was observed for four subtilases possibly representing late responsive elements of environmental stress. The kexin homologs show stronger associations with genes of transcriptional regulation context. Based on the analyses presented here and in accordance with previously characterized subtilases, we propose three main functions of subtilases: involvement in (i) control of development, (ii) protein turnover, and (iii) action as downstream components of signalling cascades. Supplemental material is available in the Plant Subtilase Database (PSDB)
(http://csbdb.mpimp-golm.mpg.de/psdb.html)
, as well as from the CSB.DB (http://csbdb.mpimp-golm.mpg.de).
Synopsis
The first complete plant genome sequence was available for Arabidopsis thaliana, a common weed. The number of genes in the Arabidopsis genome is estimated to be around 25,000. The functions of most of these gene are, however, still unknown. Many genes are grouped into gene families due to conserved sequences and predicted protein structures. In this article, the large subtilisin-like serine protease (subtilase) family of Arabidopsis is analysed. Although 56 subtilase genes have been identified in Arabidopsis, the function of only two subtilases is known. Analysis of mutants has revealed no further hints about the function of the other 54 subtilases. Here the authors present a novel approach to infer hypotheses about functions of the subtilase genes using computational analysis. Based on the analyses presented here and in accordance with previously characterized subtilases, they propose three main functions of subtilases: involvement in (i) control of development, (ii) protein degradation, and (iii) signalling. The results presented can be used to direct further analysis to elucidate functions of subtilases in plants.
Citation:Rautengarten C, Steinhauser D, Büssis D, Stintzi A, Schaller A, et al. (2005) Inferring hypotheses on functional relationships of genes: Analysis of the Arabidopsis thaliana subtilase gene family. PLoS Comput Biol 1(4): e40.
==== Body
Introduction
Subtilisin-like proteases (subtilases) are serine proteases with a catalytic triad of the three amino acids aspartate, histidine, and serine [1]. Eukaryotic subtilases belong to the S8 serine protease family (http://merops.sanger.ac.uk) and can be grouped into the pyrolysins and the kexins. Nine subtilases, the proprotein convertases, have been characterized in mammals. Of these, seven belong to the kexin subfamily, and two recently identified subtilases to the pyrolysin subfamily [2,3]. Kexin was identified as the first eukaryotic subtilase required in yeast for the processing of the precursors of α-mating factor and of killer toxin [4]. The seven mammalian kexin homologs are involved in the formation of peptide hormones, growth factors, neuropeptides, and receptor proteins from precursor polypeptides [2,3]. The two mammalian pyrolysins carry out specific cleavage and processing reactions on sterol regulatory elements, binding proteins, and pro-brain-derived neurotrophic factors, respectively [5,6]. The subtilase gene families in plants exceed in number that of mammalian subtilases by far [3]. They probably expanded to mediate a much wider range of processes. All hitherto identified plant subtilases have been grouped into the pyrolysin subfamily within the S8 serine protease family [7].
Despite the recent advances, our current understanding of subtilase functions in plants is still very limited. Currently, there is evidence for involvement of subtilases in both general protein turnover [8] and as highly specific regulation of plant development [9]. Few proteases have been purified from plant tissues and classified as subtilases based on their catalytic properties and primary structure [10–19]. Most of these enzymes are highly abundant and exhibit broad substrate specificity. Thus, a functional involvement in general protein turnover was forecasted for these abundant proteins [8,20,21]. The tomato subtilase P69 is one of several subtilases that are specifically induced following pathogen infection [22–24]. P69 processes a leucine-rich repeat cell wall protein in virus-infected tomato plants and thus is one of the very few plant subtilases for which an endogenous substrate has been identified [25]. The direct consequences of this processing event are still unknown. The P69 enzymes form a distinct subgroup among the 15 subtilases that have hitherto been cloned from tomato [26].
Forward genetics has identified subtilases as highly specific regulators of plant development. In the Arabidopsis SDD1 mutant (stomatal density and distribution 1), the pattern of stomata formation is disrupted, resulting in clustering of guard cells as well as in a dramatic increase of stomatal density [9]. The SDD1 gene is specifically expressed, and the protein is probably secreted into the apoplast [27]. Likewise, the gene disrupted in the ALE1 mutant (abnormal leaf shape 1) was cloned and found to encode a subtilase. ALE1 is required for cuticle formation and epidermal differentiation during embryo development in Arabidopsis [28]. The mutant phenotypes of SDD1 and ALE1 demonstrate that at least some subtilases carry out highly specific functions in plant development. Their modes of action in the regulation of the respective developmental processes are still unknown, but SDD1 and ALE1 may be required for the generation of peptide signals, which act non-cell autonomously to control plant development [27,28].
Despite the sequence homology–based prediction of 53 further A. thaliana subtilase (AtSBT) genes [29], there is still uncertainty about the functions of the majority of plant subtilases, including those of the model organism Arabidopsis. Here we describe results obtained with a complete set of gene knockout mutants and expression profile analysis. The main focus of our report is directed toward inference of hypotheses for functions of the so far uncharacterized Arabidopsis subtilase genes using computational analyses. We extended common classification of gene families by sequence similarity toward investigation into co-responding synchronous changes of transcript levels (co-response analyses). These generic data analysis procedures provided us with indications about the respective functional context of subtilases, which are presented here. These results demonstrate how the rapidly growing collections of gene expression profile data can be used to direct further experimental analyses to uncover gene functions.
Results/Discussion
The goal of this work was to initiate the functional characterization of the Arabidopsis subtilase gene family members. A traditional entry point was based on pairwise or multiple sequence comparisons and alignments by various algorithms [30], which provides the means for functional prediction for genes or gene products by annotation transfer from homologous sequences [31,32]. We applied this approach to identify and classify Arabidopsis subtilase genes according to their sequence homology. However, initial attempts to transfer annotation did not provide us with strong clues to draw experimentally testable hypotheses due to the lack of characterized reference genes. Moreover, verified homozygous gene knockout lines revealed no obvious phenotypic alterations and, therefore, did not support basic functional assignment. Thus, gene expression co-response analysis was performed as an alternative procedure to generate hypotheses on functional contexts of Arabidopsis subtilases.
The AtSBT Family Comprises 56 Genes
Our initial effort to identify subtilases was based on sequence comparisons with known and well-characterized Arabidopsis subtilase genes. Subtilases contain a catalytic triad (S8 domain) of the amino acid residues aspartate (Asp, D), histidine (His, H), and serine (Ser, S), as well as an asparagine (Asn, N), suggested to act as a substrate binding site. Sequence comparisons against AGI proteins (TAIR, [33]) were performed using the BLAST algorithm [34] with the S8 domain of the SDD1 amino acid sequence to identify homologous sequences. The identified sequences were evaluated for the presence of the conserved D-, H-, S-, and N-regions, and 56 subtilase encoding genes were detected (Table 1). Of the 56 genes, 55 encode proteins that contain all conserved motifs, while At5g45640 (AtSBT5.5) lacks the central Asp residue of the D-region. Hence, the subtilase family is among the largest protease gene families known in Arabidopsis. In addition to the previously identified 55 AtSBT genes [29], our analysis revealed another AtSBT containing the S8 domain, namely At4g20850 (AtSBT6.2).
Table 1 AtSBT Assignment and GenBank Accessions of the 56 Identified Arabidopsis Subtilase Genes
Beyond the sequence homology, predictions of the subcellular localization of a gene product provide additional indications for a possible functional involvement. Primary structure analysis using either TargetP [35] or PredoTar (http://genoplante-info.infobiogen.fr/predotar/predotar.html) indicated that most of the AtSBTs possess signal sequences for targeting to the secretory pathway. Six subtilases do not contain any known protein targeting motif. Three (one) family members are predicted to be targeted to mitochondria and one (three) to chloroplasts (Table 2). Experimental data for the subcellular localization of Arabidopsis subtilases are presently available only for SDD1 and for ARA12. In agreement with the predictions, they were both shown to be exported to the apoplast [19,27].
Table 2 Targeting Prediction of the 56 Arabidopsis Subtilases using Either TargetP V1.0 or PredoTar V1.03
The AtSBT Family Consists of Six Subfamilies
Relationships of the 56 Arabidopsis subtilase sequences were investigated in order to identify pairs or groups of genes that could have overlapping or similar functions according to high similarities. We performed a multiple alignment with the deduced complete amino acid sequences. The obtained neighbour-joining tree, generated from this alignment, revealed six distinct subtilase subfamilies in Arabidopsis (Figure 1). The assignment of a gene to a specific subfamily was based primarily on the position within the phylogenetic tree, as defined by the homology between the deduced full-length amino acid sequences. When a gene could not be assigned to a particular clade with a significant bootstrap value, the assignment to a certain subfamily was made by ranking BLAST search results of queries for family members against the gene. Repeating the analysis by comparing only the conserved peptidase S8 domain, we could confirm the assignments for all Arabidopsis subtilase genes into the six subfamilies. The assignments were further supported by distance matrices obtained by pairwise global alignments of the nucleic acid (Table S1) and amino acid sequences (Table S2). The protease-associated (PA) domain is supposed to determine substrate specificities of subtilases or to form protein–protein interactions [36,37]. Most proteins of the subtilase family contain a sequence region of about 120 amino acids inserted into their catalytic domain. Therefore, to uncover similar substrate specificities within the Arabidopsis subtilase family, the PA domain was used for the assignment into subfamilies. Apart from AtSBT4.1, AtSBT 6.1, and AtSBT 6.2, all Arabidopsis subtilases contain an insertion consistent with a PA domain. Apart from members of the heterogeneous subfamily 5, all subtilases were again assigned to the same subfamilies as before with only minor changes (Figure S1).
Figure 1 Bootstrapped Neighbour-Joining Tree Generated from an Alignment of the Predicted 56 AtSBT Full-Length Protein Sequences
Groups of neighbouring genes (e.g., At3g46840 and At3g46850) are distinguished by specific colours.
The general consistency of phylogenetic trees derived from the full-length and the PA domain sequences suggests that the PA domain insertion was already present in ancestral subtilases. These results are consistent with those reported by Beers et al. [29], who defined three subgroups of Arabidopsis subtilases: S8–1, S8–2, and S8–3. The AtSBT1 and AtSBT2 subfamilies assembled through our analysis are identical to the S8–2 and S8–3 groups. The large heterogeneous S8–1 group, however, was subdivided further into AtSBT families 3, 4, and 5. The AtSBT6 subfamily includes only two members: i.e., AtSBT6.1, which had not been assigned to any group by Beers et al., and AtSBT6.2, which is a previously unrecognized Arabidopsis subtilase. Both genes are characterized by a stronger similarity to the mammalian kexins and pyrolysins than to plant subtilases, whereas all other Arabidopsis SBT subfamilies do not partition with any of the known human PCs (Figure S2). According to Killer toxin processing activity detected in plant extracts, the presence of plant kexin-like subtilases has been postulated [4,38]. In contrast to the mammalian kexin homologs required for the formation of functional proteins from precursor polypeptides [2,3], the identity of the plant kexin-like subtilases responsible for the observed processing reaction is unknown.
Subtilase Families in Plants Exceed Complexity in Animals
Knowledge of phylogenetic relationships may help to unravel the basic functions of genes based on annotation transfer from orthologous sequences. BLAST searches using the peptidase S8 domain and several full-length amino acid sequences of Arabidopsis were performed via the NCBI Oryza sativa BLAST page [39] and revealed 34 non-redundant rice subtilase genes. A multiple sequence alignment of the 15 known tomato [26], the 34 identified rice, and the 56 Arabidopsis subtilases was performed to elucidate the phylogenetic relationships within the plant subtilase family. Within the obtained neighbour-joining tree, four major clusters of orthologous groups were identified that include all members of the AtSBT subfamilies 1–3 and 5, whereas AtSBT4 seems to be a subfamily specific for A. thaliana. The obtained neighbour-joining tree enabled us to identify putative orthologous pairs and groups of genes (Figure S3). However, the lack of functionally characterized orthologues in the subtilase family among the three plants species gave us no strong hints for functional annotation. Interestingly, all three plant species are characterized by significantly larger numbers of subtilases as present in animal organisms, e.g., human (nine), Caenorhabditis (four), or Drosophila (three), according to BLAST searches. The large number of Arabidopsis subtilase genes is the result of multiple duplication events (see below). Depending on the degree of functional diversification that occurred in the further evolution after the duplications, members of the gene family may have overlapping (“redundant”) or separate functions.
Chromosomal Distribution and Gene Duplications of the AtSBTs
To unravel possible redundancy, we investigated the chromosomal locations of the AtSBTs and inferred the gene duplication events that probably caused this distribution. Arabidopsis subtilase genes are present on all five chromosomes (Figure 2). The genes occur isolated or in tandem repeats, indicating that segmental and tandem duplication events may have contributed to the evolution of the Arabidopsis subtilase gene family. In contrast to the observed average of 17% on genome scale [40], 54% of AtSBT genes, belonging only to the subfamilies 3, 4, and 5, occur in tandem duplications of two up to five genes. These arrangements suggest that local duplications events also contributed to the AtSBT family expansion. Furthermore, several highly similar sequences are found on different chromosomes. Similar situations indicative of a complex evolutionary history have been observed in other Arabidopsis gene families, too [41,42]. To determine whether the formation of the AtSBT gene family is in part the result of genome duplication events, we analysed the chromosomal distribution of the AtSBTs using the Genome and Redundancy Viewer (http://mips.gsf.de/proj/thal/db/gv/rv/rv_frame.html). Analyses of the chromosomal distribution revealed that at least 18 AtSBT genes are located in previously documented segmental duplicated regions within the Arabidopsis chromosomes.
Figure 2 Physical Map of AGI at TAIR Indicating the Chromosomal Distribution of the AtSBT Family
The AtSBT genes are localized throughout the Arabidopsis genome as single genes or in tandem repeats.
Macro-scale duplication and rearrangement of chromosomes as well as micro-scale translocation and duplication are thought to be the major modes of plant genome evolution [43].
The results confirm local and segmental duplication events as the cause for expansion of the subtilase gene family in the course of the Arabidopsis genome evolution. As the two copies of a duplicated gene were initially identical and functionally redundant, the structure of the subtilase gene family poses the question: to what extent did the divergence of duplicated genes lead to the acquisition of novel and specific functions of subtilases in Arabidopsis?
Mutant Identification and Evaluation
To elucidate the functions of all Arabidopsis subtilases, T-DNA insertion mutants have been collected and analyzed for morphological traits. A total of 179 obtained T-DNA insertion lines of 55 AtSBTs have been tested by PCR with gene-specific primers for the presence of the proposed insertion, which was confirmed in 144 lines. For 44 genes, more than one verified T-DNA line is available, and for 55 AtSBT genes homozygous T-DNA insertion lines have been isolated (Table S3). Aerial organs of all homozygous lines were visually and microscopically examined at several developmental stages. Except for AtSBT1.2 (sdd1), no visible phenotypic alterations linked to the insertion were detectable. These observations suggest that either most AtSBT genes mediate specific, conditional responses, or, alternatively, that a large degree of functional redundancy exists among or within subsets of the subtilase family. Indications for the latter possibility were obtained by sequence analyses that identified groups or pairs of closely related genes (see above). To test for potential homology-based functional redundancies, we created and confirmed double knockouts and knockout/RNAi lines (see PSDB). However, none of the obtained transgenic lines exhibited any morphological phenotypic alterations. While further in-depth analysis will be necessary, including monitoring of the responses to various environmental challenges and investigation of metabolic perturbations to complete the phenotypic characterization, these observations may indicate that (partial) functional redundancy may exist even among more family members showing higher sequence divergence. In order to obtain further indications as to which pairs or groups of genes may perform similar or overlapping functions despite low degrees of sequence similarity, and what their physiological roles may be, gene expression co-response analyses were performed.
AtSBT Co-Expression and Co-Response Analyses
The increasing number of publicly available expression profiles analyzed in the frame of specific experiments enables scientists to use and to re-analyze the data for certain different questions. We investigated in such a cross-experimental approach by computational analysis of the co-expression and co-response behaviour of subtilases using 123 gene expression profiles publicly available from NASCArrays [44] and 192 profiles of the AtGenExpress developmental series [45]. The expression profile data were generated using the Ath1 gene chip technology platform (Affymetrix), which contains specific oligonucleotides for 52 of the 56 annotated AtSBT genes. We focused initially only on the AtSBT genes to compare the expression within the subtilase family. This analysis was (i) first, performed using the qualitative attributes “present,” “marginal,” and “absent” of the array technology platform, (ii) then extended to quantitative values of (relative) expression levels, and (iii) third, widened to include all other genes, allowing us to assign subtilases to defined functional classes based on their co-response behaviour with functional classified non-subtilase genes. The goal of our computational expression analysis was to infer experimentally testable hypotheses regarding the (i) functional interplay of AtSBTs and (ii) the functional contexts in which theses gene may be embedded.
Ubiquitous and Conditional Expression of AtSBTs Revealed by Co-Expression Analysis
Our first computational investigation regarding the expression behaviour of subtilase genes focused on the grouping of AtSBTs according their coherent expression under identical experimental conditions. Coherent gene expression identifies either ubiquitous or conditional expression and allows first insight into a possible functional interplay of genes. To investigate the co-expression of AtSBTs, we converted the detection calls into qualitative Boolean values: (i) absent and marginal detection calls were set to null and (ii) present calls to one. Pairwise distances among all genes were computed using the S9 index via bootstrap analyses [46] with 999 numbers of bootstrapped Boolean matrices for each dataset (see Materials and Methods). The corresponding distance matrix was subjected to hierarchical cluster analysis (HCA) of the genes. Cluster trees drawn on the basis of bootstrap and non-bootstrap analyses revealed perfect agreement, i.e., the genes were assigned into the same cluster with same tree sorting. Validity and statistical significance of the clusters and the cluster tree structure were supported by bootstrap support values drawn on the basis of the resulting consensus cluster tree (see Materials and Methods). As a result of this analysis, we identified two most distantly related AtSBT gene clusters (Figure 3, Table 3, PSDB). Gene cluster I contained 18 (32%) of the 54 represented AtSBT genes and showed the following subfamily representation: AtSBT1: seven (78%), AtSBT2: five (83%), AtSBT3: zero (0%), AtSBT4: one (7%), AtSBT5: three (50%), AtSBT6: two (100%). In contrast, gene cluster II contained 35 (62%) AtSBT genes, which all belong to the subfamilies AtSBT1, AtSBT2, AtSBT3, AtSBT4, and AtSBT5. Whereas cluster I mainly represents ubiquitously expressed genes with some expressed at high levels, cluster II primarily contains genes with specific expression pattern and/or low expression levels. To confirm the obtained co-expression behaviour of AtSBTs and to test for biological relevance underlying the statistically significant clusters, we investigated their tissue-specific expression using semi-quantitative RT-PCR analyses. The obtained organ-specific expression patterns of the analyzed genes revealed ubiquitous expression for cluster I genes (Figure 3). The genes assigned to cluster II, on the other hand, exhibited expression primarily in one organ or in a subset of the analyzed organs. For some genes assigned to cluster II, namely AtSBT1.2, 1.9, 3.5, 3.6, and 3.12, we confirmed expression pattern for most of the analyzed organs. According to the results obtained by both analyses, we concluded that genes of cluster I are constitutively expressed, both in terms of organ specificity as well as according to various conditions. In contrast, genes of cluster II mainly show specific expression patterns (Figure 3).
Figure 3 Bootstrapped Consensus Cluster Tree from Converted Detection Call Matrix of Affymetrix (Ath1) Microarray Experiments into Boolean Values (AtGenExpress Developmental Series)
Cluster I covers ubiquitously expressed genes, whereas Cluster II mainly represents lowly or specifically expressed genes. These results were validated independently by semi-quantitative RT-PCR analysis (shown in the right panel). CL, cauline leaves; dS, dry seed; F, flower; R, root; RL, rosette leaves; Sd, seedling; Sq, siliques; St, inflorescence stem.
Table 3 AtSBT Family Member Distribution within the Main Cluster I and II
Transcriptional Interrelation among AtSBTs Revealed by Co-Response Analyses
While through the (qualitative) co-expression the global activity profiles of the AtSBT genes were revealed, the (quantitative) co-response analysis was performed to identify pairs or groups of AtSBT genes that show similar transcript changes among a multi-conditional set of expression. This analysis was carried out to test for overlapping and possible redundant functional interplays of AtSBTs regarding their expression behaviour. For our subsequent analyses, we implicitly make the assumption that common transcriptional control of genes is reflected in co-responding, simultaneous changes in transcript levels [47]. A necessary prerequisite for a co-response analysis is a considerable variation of gene expression levels across the datasets used. Furthermore, valid measures of expression, i.e., values above the detection limit, have to be available for the genes in question in most, ideally all, profiles. For the analysis, three multi-conditional gene expression data matrices (replicates) were assembled, each consisting of one of three replicates of approximately 50 expression profiles. The 50 selected datasets were composed approximately equally by a range of examined experimental condition (out of a total of 123 replicated experiments). These matrices were maximised for the diversity of the represented experimental conditions. Each of them covered series of valid gene expression values of approximately 10,000 genes, including 12 AtSBT genes, with valid measured transcript levels.
Our numerical approach to detect transcript co-responses is based on the non-parametric Spearman's rank order correlation (rs), which is a robust estimation of correlation. For bias estimation, as well as for a more exact approximation of the statistical probability, we performed iterative computation of rs based on bootstrap analysis. A test of homogeneity was applied to compare the co-responses derived from the three data matrices and revealed no significant differences among the pairwise transcript co-responses. As the test of homogeneity can detect only large differences among pairwise transcript co-responses derived from different data matrices, we applied in addition the mantel test, performed as non-parametric Spearman correlation of matrices. The mantel test was used to estimate the association between the three independent data matrices describing the same set of entities. It tests whether the association between the matrices is stronger than a random association, i.e., whether similar transcript co-responses and similar biologically relevant information is coherent in the different matrices. This analysis showed highly significant correlations (p << 0.001) in the range of 0.87 ≤ rs ≤ 0.90, with an average of 0.89 ± 0.02 among the matrices. Both statistical tests revealed that similar information about transcript co-responses could be deduced from the matrices. Therefore, the common transcript co-responses could be computed and used for HCA. Figure 4A shows a cluster tree drawn on the basis of the common co-responses for the three matrices. Spearman correlations (−1 < = rs < = 1) were converted into distance measures (d) by the simple transformation d = 1−rs. Genes that showed opposing changes of transcript levels (with most negative co-responses) thus were displayed with largest distances, while the most highly correlated genes had smallest distances. The 12 represented AtSBT genes were grouped into three well-separated clusters: (i) with AtSBT2.5, AtSBT1.4, AtSBT1.7, AtSBT1.6, and AtSBT5.6, (ii) with AtSBT2.1, AtSBT1.8, AtSBT1.5, and AtSBT1.3, and (iii) with AtSBT6.1, AtSBT4.14, and AtSBT6.2. The joints were at relatively large heights and reflected that the corresponding changes in transcript levels were not identical but similar among pairs and groups of AtSBT genes. For further analysis, we visualized as a network significant, Bonferroni corrected [48] correlations among the AtSBT genes using the Pajek software [49] (Figure 4B). In conjunction with the cluster tree drawn on the basis of the common co-responses (Figure 4B), the obtained AtSBT network revealed two AtSBT cliques, where each gene member showed significant correlation to the other members. Clique I covered AtSBT1.4, AtSBT1.6, AtSBT1.7, AtSBT5.6, and AtSBT2.5, whereas the clique II included the genes AtSBT1.3, AtSBT1.5, AtSBT2.1, and AtSBT2.5. The subtilase AtSBT2.5 is shared between both cliques and represents a “hub” within the AtSBT network, which shows significant connections to all genes of the two main cliques and interconnects both AtSBT cliques (Figure 4). The average co-response of AtSBT2.5 to both cliques was 0.47 ± 0.08. Exclusion of AtSBT2.5 revealed an average co-response of 0.69 ± 0.08 within clique I and of 0.43 ± 0.01 within clique II. AtSBT1.8 is positively correlated with AtSBT2.5, but this gene shows less connectivity to the two cliques.
Figure 4 Transcript-Level Correlations of Ubiquitously Expressed Subtilase Genes in Multi-Conditional 22k Affymetrix Expression Profiles
(A) shows a cluster tree of the of the transcript level correlations of ubiquitously expressed subtilase genes in multi-conditional 22k Affymetrix expression profiles. Correlations were determined by computing the Spearman's rank order correlation. (B) shows a network of all significantly positive associated AtSBTs according to the average Spearman correlation.
The statistical analyses revealed significant co-responses among AtSBT genes, but the causality of the interrelations remains to be shown. Non-parametric Kendall's tau (τ) correlation of E. coli operon genes controlled by common cis-elements revealed a co-response distribution over a broad range [47]. Considering the relationship of Spearman's rs and Kendall's τ (rs~ 3/2 τ), the co-responses among AtSBT genes of clique I are in the upper range of these distributions. Therefore, a biological relevance of the observed co-response network can be assumed. In conjunction with the results of semi-quantitative RT-PCR and the co-expression analysis (see above and Figure 3), we conclude that the genes of clique I are ubiquitously but not constitutively expressed and that they respond to similar cues. The revealed associations and the central positions of AtSBT2.5 and AtSBT1.8 in the network suggest that both genes might be involved in the same functional context and may have (partially) overlapping roles. However, the similarities of the amino acid (32.9%) and the nucleic acid sequences (50.0%) of these two genes are not higher than their homologies to other AtSBT genes (average 34.7%/50.5%; see PSDB). In contrast, AtSBT2.5 is highly related to AtSBT2.6 (aa: 88.1%; nt: 83.7%), both are ubiquitously expressed, and have probably evolved from a sequential duplication. Redundancy of function that might be assumed according to the close evolutionary relationship was not supported, as a verified double homozygous T-DNA insertion line did not show any visible mutant phenotype despite their similar expression pattern (Figure 3). Similarly, AtSBT5.6 is highly related to AtSBT5.5 at the sequence level (aa: 62.3%; nt: 68.0%), but only AtSBT5.6 is a member of clique I. Sequence similarities between AtSBT5.6 and other members of clique I (average aa: 40.5%; average nt: 52.9%) are not notably higher than other AtSBT genes (average aa: 36.9%; average nt: 51.3%). In contrast, AtSBT1.4, AtSBT1.6, and AtSBT1.7 represent an example of evolutionary related genes with higher-than-average homology on the amino acid (46.8%–54.3%) and nucleic acid (57.7%–59.7%) level that are members of the clique I and show significant co-regulation. Nevertheless, AtSBTs with even higher sequence homology but lower co-response are present in subfamily 1. The co-response analysis of the AtSBT gene family thus revealed potential functional relationships, which in some cases clearly contradicted the predictions made on the basis of sequence analysis. In conclusion, we suggest that even minor differences in sequence similarity may confer functional divergence and that functional redundancy within the Arabidopsis subtilase family may be better revealed by transcriptional co-response analysis than by high sequence similarity. It is very likely that a few amino acid changes could alter the substrate specificity of a protease. A striking example of the consequences of a single amino acid change on the properties of an enzyme is provided by the stilbene synthases [50].
Co-Response-Based Transcriptional Neighbourhood Search of AtSBTs
As a third step, we extended our co-response analyses to the characterization of the co-responses of AtSBT genes with all other genes represented in the underlying data matrices. This was performed to identify sets of co-regulated genes that are assigned to certain functional categories and may provide information on the functional context of individual or groups of AtSBT genes.
The degree of transcript co-responses may be influenced by the selection of the experiments used for generating the (multi-)conditional data matrices, and predictions based on nearest neighbours may be of equivocal nature. However, we assumed that the enrichment of transcriptionally correlated genes of a certain functional category should be a more robust marker of the functional context of a gene of interest. To obtain such indications for the AtSBT genes, we selected the top two percent of the strongest positive as well as negative correlated genes to each AtSBT gene. We computed the enrichment by adding up relative impacts (RIs) of the genes assigned to particular functional categories, where the gene-specific RI was defined as the reciprocal of the number of assignments of a gene to different categories. As reference, we calculated the enrichment as mentioned above over all genes represented in the underlying data matrices.
Applying the G-test of independence, which tests hypotheses about frequencies, for the positive best two percent correlated genes (Figure 5A) revealed that genes belonging to the category “unclassified” were significantly enriched (p << 0.001) for each of the 12 AtSBT genes, with an average of 1.78-fold. A significant (p < 0.05) enrichment of genes assigned to “metabolism” and “energy” was observed for AtSBT5.6. For AtSBT1.6, a member of the clique I (Figure 4), we detected a tendency (p < 0.1) of enrichment for “metabolism.” For AtSBT1.5, a member of the clique II (Figure 4), a significant enrichment for “control of cellular organisation” was observed. To categorise AtSBT genes according to their neighbourhood, we normalized each category-specific sum of RIs and expressed it as the fraction of the sum of all RIs over all categories. The co-responding matrix was subsequently used for hierarchical cluster analyses on the basis of the functional context in the neighbourhood by computing Euclidean distances. According to the obtained cluster tree for positive associated neighbourhood (Figure 5A), we suggested a similar functional context for AtSBT2.5, the major hub connecting cliques I and II (Figure 4B) and AtSBT1.8 that showed lower connectivity to the two cliques. Interestingly, analysis based on the two percent of strongly negative associated genes (Figure 5B) revealed different neighbourhoods for the two genes. According to these results, and consistent with the results of the co-expression (Figure 3) and co-response (Figure 4) analyses, we suggest that AtSBT2.5 and AtSBT1.8 have overlapping but not identical functions. The hub AtSBT2.5 and AtSBT1.8 are characterized by an enrichment of positively correlated genes assigned to “cellular communication/signal transduction mechanism” as well as “cellular organization,” which are ranked at positions 2 and 3.
Figure 5 Result of the Co-Response-Based Transcriptional Neighbourhood Search for 12 Ubiquitously Expressed Subtilase Genes
The best 2% of positively correlated (A) and of negatively correlated (B) genes were selected and used to determine functional category representations. The upper chart represents the cluster tree resulting from our HCA analysis based on conversion of the enrichment of genes of particular functional categories into the Euclidean distances. In the lower chart, vertically stacked bar plots illustrate the distribution of functional categories of correlated gene for each of the 12 AtSBTs. For comparison, the functional category distribution of the genes represented in the underlying data matrices is shown to the left of each display.
Moreover, for the genes AtSBT1.4 and AtSBT1.7, as well as for AtSBT1.6 and AtSBT5.6, the members of clique I (Figure 4B), we observed early joining, according to the representation of functional classes by both the strongly positive and, with exception of AtSBT1.4, the strongly negative associated genes. According to a significant enrichment of genes assigned to the functional category “metabolism” (Figure 5A, PSDB), we suggest that these genes are embedded in the functional context of metabolism. These four genes were also correlated in expression with genes enriched for functions in “cell rescue, defence and virulence” and in “transport facilitation,” which are ranked at positions 3 and 4 (PSDB). The correlated behaviour and similar functional neighbourhoods of this set of AtSBTs hints to an involvement within the physiological context of pathogen response and/or general stress-related responses. The indications obtained for the functional contexts of these two sets of AtSBT genes leads us to suggest that AtSBT2.5 and AtSBT1.8 may be involved in sensing mechanisms, or might be early responsive factors. On the other hand, AtSBT1.4, AtSBT1.6, AtSBT1.7, and AtSBT5.6 may be related to more specific downstream processes. Such involvement in similar or identical processes may lead to similar transcriptional co-responses among genes as it was observed for the above-mentioned subtilases (Figure 4). Moreover, the HCA analyses of the neighbourhoods (Figure 5A) agreed with the tree and network drawn on the basis of the transcriptional interrelation among the ubiquitously expressed subtilases (Figure 4). The experimental verification of this hypothesis will be one of the goals of our continuing functional genomics project on the characterization of plant subtilases.
The PSDB
Our research aims at the functional characterization of the subtilases in Arabidopsis. To achieve this goal, an international consortium of five European and US partners (The Arabidopsis Subtilase Consortium; http://csbdb.mpimp-golm.mpg.de/psdb.html) was established. The multiple levels of comprehensive data accumulated in this project by us and other groups within The Arabidopsis Subtilase Consortium need a specialized Web interface to store and distribute data related to plant subtilases. According to these needs, we established the PSDB, which is an associated database of the Comprehensive Systems-Biology Database (http://csbdb.mpimp-golm.mpg.de). PSDB contains confirmed results of replicated experiments related to plant (Arabidopsis) subtilase genes and allows open access to the science community. PSDB will be regularly updated with results of co-response analyses, performed on the increasing number of publicly available gene expression profiles. Furthermore, validated information of tissue-specific expression patterns of AtSBT genes, cellular localisation of encoded proteins, and phenotype information of the mutants and transgenic plants will be displayed and regularly updated. Further information and supplemental material will be available at PSDB (http://csbdb.mpimp-golm.mpg.de/psdb.html).
Materials and Methods
Sequence analysis.
Nucleic acid and amino acid sequences were retrieved by searching public databases with the BLAST algorithm [34] at TAIR
(http://www.arabidopsis.org/), TIGR
(http://www.tigr.org/),
NCBI (http://www.ncbi.nlm.nih.gov),
and MIPS (http://mips.gsf.de/).
Subcellular localization was predicted using either TargetP V1.0 [35] (http://www.cbs.dtu.dk/services/TargetP/) or PredoTar V1.03 (http://genoplante-info.infobiogen.fr/predotar/predotar.html). The deduced amino acid sequences were aligned using the CLUSTALX program [51] with the default parameter settings and manually improved with respect to all known conserved subtilase motifs. The phylogenetic tree was obtained with the neighbour-joining method with bootstrap values generated from 1,000 bootstrap samples and visualized by using the TreeView application [52].
Only bootstrap values higher than 70% were considered to be significant [53]. Bootstrap values lower than 60% are not shown.
Plant material and growth conditions.
Seeds of A. thaliana accessions Columbia-0 (Col-0), Wassilewskija, and the appropriate T-DNA mutant lines were surface-sterilized and germinated on half-concentrated Murashige and Skoog medium (M02 555 [pH 5.6], Duchefa, Haarlem, The Netherlands), supplemented with 1% sucrose, and solidified with 0.7% agar under a 16-h day (140 μmol·m−2·s−1, 22 °C)/8-h night (22 °C) regime. After 2 wk, plants were transferred to standard soil (Einheitserde GS90; Gebrueder Patzer, Sinntal-Jossa, Germany) and further grown in a growth chamber under a long-day light regime (16 h of fluorescent light [120 μmol·m−2·s−1] at 20 °C and 60% relative humidity/8 h of dark at 16 °C and 75% relative humidity).
Mutant collection, confirmation, and phenotypic analysis.
T-DNA insertion mutants were retrieved from the SIGnAL [54], the GABI-Kat [55], the Genoplante FST/FLAGdb [56], the SAIL collection (Syngenta Biotechnology, Research Triangle Park, North Carolina, United States), and the University of Wisconsin Knockout facility. Genomic DNA was isolated using the DNeasy 96 Plant Kit (Qiagen, Hilden, Germany) and subsequently used for PCR analysis. The T-DNA insertion lines were screened for the appropriate insert using the required T-DNA and a gene-specific primer. Gene-specific flanking primers were used to confirm homozygosity. Primer sequences are available at PSDB. Homozygous insertion lines were evaluated for phenotypic alterations at the following developmental stages [57]: 1.03 for seedlings grown on synthetic media, 3.9 for rosette leaves, and 6.9 for inflorescence stem, cauline leaves, flower, and siliques. Plants were examined for leaf number, shape and size, epidermal constitution with respect to trichome and guard cell number and distribution, flowering time, and flower silique and seed morphology.
Data source and pre-processing.
123 publicly available expression profiles from 22 experiments were obtained from NASCArrays (http://affymetrix.arabidopsis.info/, October 2003 [44]) and used for the generation of the data matrices nasc0271–0273. Additionally, 192 wildtype microarray profiles of 64 different tissues or developmental stages from the AtGenExpress developmental series generated at the Weigel laboratory have been used (http://www.arabidopsis.org/ [45]). The profiles were obtained through use of the Affymetrix Ath1 chip technology (Affymetrix, Santa Clara, California, United States), and results were scale normalized to TGT 100. The number of Present and Marginal calls (according to the MAS 5.0 algorithm) was calculated for each profile.
Transcript co-responses were retrieved from CSB.DB [58] for data matrix nasc0271. Co-responses for the additional matrices nasc0272 and nasc0273 were computed within this work (see below). In the majority of cases, two or three profiles per experiment with the highest numbers of Present and Marginal calls were selected for nasc0271. In analogy, nasc0272 and nasc0273 were generated from profiles per experiments ranked second and third according to the numbers of Present and Marginal calls. Thus, each of the data matrices comprised approximately 50 out of 123 profiles approximately equally representing the 22 underlying experiments, with approximately 10,000 out of more than 22,000 genes: nasc0271: 51 experiments with 9,694 genes, nasc0272: 51 experiments with 8,927 genes, and nasc0273: 49 experiments with 8,691 genes, each well measured in at least 85% of the underlying expression profiles. Transcript co-responses were computed on data matrices with log base 2 transformed and range-normalised transcript intensities for each gene.
Co-expression analysis.
For co-expression analyses, the detection calls were converted into Boolean values. The profiles of the developmental series were separated into three data matrices according to the number of replicated experiments, whereas the NASC arrays are combined in one matrix. The numerical value null was assigned to absent and marginal calls, whereas present calls were set to one. Pairwise distances among entities, i.e., genes, of the Boolean matrix were computed using the S9 index via bootstrap analyses with 999 numbers of bootstrapped Boolean data matrices [46]. Each of the generated 999 pairwise distance matrices were subsequently used for HCA [59]. The computation was executed with the statistical software environment R2.1.0 [60]. HCA was performed as unweighted average linkage clustering algorithm. The resulting hierarchical cluster trees were converted into newick tree format with the function “hclust2phylog” of “ade4” package [60] implemented in the R software [61]. The resulting consensus tree was computed with the program “consensus” of the Phylogeny Inference Package [62] of all bootstrapped cluster trees, i.e., 2,997 trees represent the basis of the consensus tree of the developmental dataset.
Semi-quantitative RT-PCR expression analysis.
Samples of the appropriate organs from Arabidopsis Col-0 plants were harvested at the stages used for mutant screening (see above). Total RNA was isolated with TRIzol reagent (Invitrogen, Carlsbad, California, United States) according to the manufacturer's protocol. 1 μg of total RNA was pre-treated with DNaseI (Ambion, Austin, Texas, United States) and reverse transcribed with SuperScriptII reverse trascriptase (Invitrogen) and d(T)15. The cDNA reaction was diluted 1:5 with water, and 5 μl of the diluted cDNA was used as template for PCR analysis applying Tit Taq PCR Enzyme System (BD Biosciences, Palo Alto, California, United States) according to the manufacturer's protocol. In general, due to the low abundance of subtilase transcripts, 40 cycles were performed with a PTC-200 thermal cycler (MJ Research, Waltham, Massachusetts, United States). Primer sequences and the size of cDNA and genomic amplicons are available via PSDB. AtACT2 was used as internal standard [63].
Co-response analyses.
A general bivariate normality cannot be assumed for each gene pair, respectively analysed with the Cramer-test [64]; transcript co-response analyses were performed by computation of non-parametric Spearman's rank order correlation (rs) [65]. Co-response analysis among AtSBT genes were calculated by non-parametric bootstrap analyses with 2,000 numbers of bootstrap samples [66]. Mantel test and test on homogeneity [65] were used to compare and compute the common correlations among different co-response matrices. Testing homogeneity was performed using Microsoft Excel. The mantel test, computed as non-parametric Spearman correlation of (dis-)similarity matrices, was executed in R1.8.1. [61]. Common Spearman correlations and joint probabilities among the matrices were calculated as recommended [65]. In order to generate normalised distance matrices, correlations were converted into distance ranges [65]. Negative Spearman correlations were assigned to be most distant and converted into the largest distances: distance = 1 – rs. Normalization of the obtained distance matrix was done by dividing distances with the obtained maximum. HCA was performed as mentioned above. Visualization of significant associations was done with the software Pajek [49]. The multiple comparison performed required the adjustment of α to accept significant associations, which was done by application of the Bonferroni correction α′ = α/k. The corrected α′ for 12 comparisons was 0.00416.
Transcriptional neighbourhood search.
The assignment of gene products to functional categories was retrieved from MAtDB (December 2003 [67]). The functional categorization is tree-like, and each category is further divided into subcategories. We used only the highest branch for each category, which included 99 categories, 29 assigned with a category name. Categories without category name were merged into the “undefined” category, and the categories 40, 43, 45, and 47 were merged into the class “localization.” Genes without assignment or with unclear classification were treated as “unclassified.” Genes assigned to more than seven categories, which represents 5% of the whole annotation, were also treated as “unclassified.” The RI of a gene with multiple assignments (nassign) onto each category was defined as ri = 1/nassign.
The transcriptional neighbourhood search was performed as follows: For each AtSBT gene, the best 2% of positively and negatively correlated genes to each represented AtSBT gene were extracted and grouped according to their assigned functional category. For calculation of the enrichment of functional categories, the sum of all RIs for each category was computed. As reference against which the enrichment/de-enrichment was determined, the sum of all RIs for each category over all represented genes was used. Significance of the observed category enrichments for each of the AtSBT was calculated for each functional category by G-test of independence [64].
Supporting Information
Figure S1 Bootstrapped Neighbour-Joining Tree Generated from an Alignment of the Predicted PA Domain using Clustal X 1.81
The tree was displayed by TreeView and edited manually. Notice that out of 56 AtSBTs, three (6.1, 6.2, and 4.1) do not contain a PA domain.
(159 KB PDF)
Click here for additional data file.
Figure S2 Bootstrapped Neighbour-Joining Tree with 1,000 Bootstrap Replicates Generated from an Alignment of the Full-Length Protein Sequences of the AtSBT1, 2, and 6 Subfamily Members, Yeast Kex2p, and the Human Prohormone Convertases (PCs, Furin, SK1) using Clustal X 1.81
The tree was displayed by TreeView and edited manually.
(83 KB PDF)
Click here for additional data file.
Figure S3 Phylogenetic Tree of Plant Arabidopsis, Tomato, and Rice Subtilisin-Like Serine Proteases
The Neighbour-Joining tree was generated from an alignment of the 56 AtSBT, 14 tomato (blue shaded), and 34 identified rice (black font) full-length protein sequences. Branch lengths are proportional to number of amino acid substitutions. Four major clusters of orthologous groups (MCOGs) were identified that included all members of the Arabidopsis subfamilies 1, 2, 3, and 5. The Arabidopsis subfamily 4 seems to be specific for this plant species, whereas with the exception of TMP, all analyzed tomato subtilases belong to the MCOG2.
(576 KB PDF)
Click here for additional data file.
Table S1 Distance Matrix Obtained by a Pairwise Global Alignment of the 56 AtSBT Nucleic Acid Sequences
(46 KB XLS)
Click here for additional data file.
Table S2 Distance Matrix Obtained by a Pairwise Global Alignment of the 56 AtSBT Full-Length Amino Acid Sequences
(46 KB XLS)
Click here for additional data file.
Table S3 Table of the T-DNA Insertion Mutants Collected and Tested by PCR with Gene-Specific Primers for the Presence of the Proposed Insertion and Analyzed for Morphological Traits
(11 KB PDF)
Click here for additional data file.
We thank the Salk Institute Genomic Analysis Laboratory, INRA, GABI-Kat, and Syngenta Biotechnology for providing the sequence-indexed Arabidopsis T-DNA insertion mutants. Further, we acknowledge NASCArrays for the establishment of a publicly accessible repository for microarray data as well as all scientists who submitted transcript profile data to this database and thereby enabled us to perform the presented investigations. Furthermore, the comments from the three anonymous reviewers are gratefully acknowledged. This work was supported by a grant of the Deutsche Forschungsgemeinschaft (DFG) to TA (AL387/5–1; AL387/5–2) and was carried out in the frame of the Arabidopsis Functional Genomics Network (AFGN) program.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. C. Rautengarten, D. Steinhauser, D. Büssis, A. Stintzi, A. Schaller, J. Kopka, and T. Altmann conceived and designed the experiments. C. Rautengarten, D. Steinhauser, D. Büssis, and A. Stintzi performed the experiments and analyzed the data. C. Rautengarten, D. Steinhauser, and A. Schaller contributed reagents/materials/analysis tools. C. Rautengarten, D. Steinhauser, D. Büssis, A. Schaller, J. Kopka, and T. Altmann wrote the paper.
Abbreviations
AtSBT
Arabidopsis thaliana subtilase
HCAhierarchical cluster analysis
PAprotease-associated
PSDBPlant Subtilase Database
RIrelative impact
==== Refs
References
Dodson G Wlodawer A 1998 Catalytic triads and their relatives Trends Biochem Sci 23 347 352 9787641
Barr PJ 1991 Mammalian subtilisins: The long-sought dibasic processing endoproteases Cell 66 1 3 2070411
Seidah NG Chrétien M 1999 Proprotein and prohormone convertases: A family of subtilases generating diverse bioactive polypeptides Brain Res 845 45 62
Fuller RS Brake A Thorner J 1989 Yeast prohormone processing enzyme (KEX2 gene product) is a Ca2+-dependent serine protease Proc Natl Acad Sci U S A 86 1434 1438 2646633
Sakai J Rawson RB Espenhade PJ Cheng D Seegmiller AC 1998 Molecular identification of a sterol-regulated luminal protease that cleaves SREBSs and controls lipid composition in animal cells Mol Cell 2 505 515 9809072
Seidah NG Mowla SJ Hamelin J Mamarbachi AM Benjannet S 1999 Mammalian subtilisin/kexin isozyme SKI-1: A widely expressed proprotein convertase with a unique cleavage specificity and cellular localization Proc Natl Acad Sci U S A 16 1321 1326
Siezen RJ Leunissen JAM 1997 Subtilases: The superfamily of subtilisin-like serine proteases Protein Science 6 501 523 9070434
Bogacheva AM 1999 Plant subtilisins Biochemistry (Mosc) 3 287 293
Berger D Altmann T 2000 A subtilisin-like serine protease involved in the regulation of stomatal density and distribution in Arabidopsis thaliana
Genes Dev 14 1119 1131 10809670
Rudenskaya GN Bogdanova EA Revina LP Golovkin BN Stepanov VM 1995 Macluralisin—A serine proteinase from fruits of Maclura pomifera (Raf.) Schneid Planta 196 174 179 7767235
Rudenskaya GN Bogacheva AM Preusser A Kuznetsova AV Dunaevsky YE 1998 Taraxalisin—A serin proteinase from dandelion Taraxacum officinale Webb s.l FEBS Lett 437 237 240 9824298
Bogacheva AM Rudenskaya GN Dunaevsky YE Chestuhina GG Golovkin BN 2001 New subtilisin-like collagenase from leaves of common plantain Biochimie 83 481 486 11506892
Terp N Thomsen KK Svendsen I Davy A Simpson DJ 2000 Purification and characterization of hordolisin, a subtilisin-like serine endoprotease from barley J Plant Physiol 156 468 476
Fontanini D Jones BL 2002 SEP-1—A subtilisin-like serine endopeptidase from germinated seeds of Hordeum vulgare L. cv. Morex Planta 215 885 893 12355148
Dyer JM Chapital DC Lax AR Pepperman AB 1999 Identification of a subtilisin-like protease in seeds of developing tung fruits J Plant Physiol 155 802 805
Popovic T Puizdar V Brzin J 2002 A novel subtilase from common bean leaves FEBS Lett 23 163 168
Beilinson V Moskalenko OV Livingstone DS Reverdatto SV Jung R 2002 Two subtilisin-like proteases from soybean Physiol Plant 115 585 597 12121465
Boyd PM Barnaby N Tan-Wilson A Wilson KA 2002 Cleavage specificity of the subtilisin-like protease C1 from soybean Biochim Biophys Acta 29 269 282
Hamilton JM Simpson DJ Hyman SC Ndimba BK Slabas AR 2003 Ara12 subtilisin-like protease from Arabidopsis thaliana: Purification, substrate specificity, and tissue localization Biochem J 15 57 67
Kaneda M Tominaga N 1975 Isolation and characterization of a proteinase from the sarcocarp of melonfruit J Biochem (Tokyo) 78 1287 1296 5423
Yamagata H Masuzawa T Nagaoka Y Ohnishi T Iwasaki T 1994 Cucumisin, a serine protease from melon fruits, shares structural homology with Subtilisin and is generated from a large precursor J Biol Chem 269 32725 32731 7806492
Vera P Conejero V 1988 Pathogenesis-related proteins of tomato, P-69 as an alcaline endoprotease Plant Physiol 87 58 63 16666127
Tornero P Conejero V Vera P 1996a Primary structure and expression of a pathogen-induced protease (PR-69) in tomato plants: Similarity of functional domains to subtilisin-like endoproteases Proc Natl Acad Sci U S A 93 6332 6337 8692815
Jordá L Coego A Conejero V Vera P 1999 A genomic cluster containing four differentially regulated subtilisin-like processing protease genes is in tomato plants J Biol Chem 274 2360 2365 9891003
Tornero P Mayda M Gomez MD Canas L Conejero V 1996 Characterization of LRP, a leucine-rich repeat (LRR) protein from tomato plants that is processed during pathogenesis Plant J 10 315 330 8771787
Meichtry J Amrhein N Schaller A 1999 Characterization of the subtilase gene family in tomato (Lycopersicon esculentum Mill.) Plant Mol Biol 39 749 760 10350089
von Groll U Berger D Altmann T 2002 The subtilisin-like serine protease SDD1 mediates cell-to-cell signaling during Arabidopsis stomatal development Plant Cell 14 1527 1539 12119372
Tanaka H Onouchi H Kondo M Hara-Nishimura I Nishimura M 2001 A subtilisin-like serine protease is required for epidermal surface formation in Arabidopsis embryos and juvenile plants Development 128 4681 4689 11731449
Beers EP Jones AM Dickerman AW 2003 The S8 serine, C1A cysteine and A1 aspartic protease families in Arabidopsis Phytochemistry 65 43 58
Hodgman TC 2000 A historical perspective on gene/protein functional assignment Bioinformatics 16 10 15 10812472
McGeoch DJ Davison AJ 1986 Alphaherpesviruses possess a gene homologous to the protein kinase gene family of eukaryotes and retroviruses Nucleic Acids Res 14 1765 1777 3005981
Bork P Gibson TJ 1996 Applying motif and profile searches Methods Enzymol 266 162 184 8743684
Rhee SY Beavis W Berardini TZ Chen G Dixon D 2003 The Arabidopsis Information Resource (TAIR): A model organism database providing a centralized, curated gateway to Arabidopsis biology, research materials and community Nucleic Acids Res 31 224 228 12519987
Altschul SF Gish W Miller W Myers EW Lipman DJ 1990 Basic local alignment search tool J Mol Biol 215 403 410 2231712
Emanuelsson O Nielsen H Brunak S von Heijne G 2000 Predicting subcellular localization of proteins based on their N-terminal amino acid sequence J Mol Biol 300 1005 1016 10891285
Mahon P Bateman A 2000 The PA domain: A protease-associated domain Protein Sci 9 1930 1934 11106166
Luo X Hofmann K 2001 The protease-associated domain: A homology domain associated with multiple classes of proteases Trends Biochem Sci 26 147 148 11246007
Kinal H Park CM Berry JO Koltin Y Bruenn JA 1995 Processing and secretion of a virally encoded antifungal toxin in transgenic tobacco plants: Evidence for a Kex2p pathway in plants Plant Cell 7 677 688 7647561
Wheeler DL Church DM Edgar R Federhen S Helmberg W 2004 Database resources of the National Center for Biotechnology Information: Update Nucleic Acids Res 32 35 40 14704341
The Arabidopsis Genome Initiative 2000 Analysis of the genome sequence of the flowering plant Arabidopsis thaliana
Nature 408 796 815 11130711
Shiu SH Bleecker AB 2001 Receptor-like kinases from Arabidopsis form a monophyletic gene family related to animal receptor kinases Proc Natl Acad Sci U S A 98 10763 10768 11526204
Mladek C Guger K Hauser MT 2003 Identification and characterization of the ARIADNE gene family in Arabidopsis. A group of putative E3 ligases Plant Physiol 131 27 40 12529512
Bancroft I 2000 Insights into the structural and functional evolution of plant genomes afforded by the nucleotide sequences of chromosomes 2 and 4 of Arabidopsis thaliana
Yeast 17 1 5 10797596
Craigon DJ James N Okyere J Higgins J Jotham J 2004 NASCArrays: A repository for microarray data generated by NASC's transcriptomics service Nucleic Acid Res 32 D575 D577 14681484
Schmid M Davison TS Henz SR Pape UJ Demar M 2005 A gene expression map of Arabidopsis thaliana development Nat Genet 37 501 506 15806101
Gower JC Legendre P 1986 Metric and Euclidean properties of dissimilarity coefficients J Classification 3 5 48
Steinhauser D Junker BH Luedemann A Selbig J Kopka J 2004 Hypothesis-driven approach to predict transcriptional units from gene expression data Bioinformatics 20 1928 1939 15044239
Bonferroni CE 1935 Il calcolo delle assicurazioni su gruppi di teste Studi in onore del Professore Salvatore Ortu Carboni Rome. pp 13 60
Batagelj V Mrvar A 2003 Pajek—Analysis and visualization of large networks Jünger M Mutzel P Graph drawing software Berlin Springer 77 103
Suh DY Fukuma K Kagami J Yamazaki Y Shibuya M 2000 Identification of amino acid residues important in the cyclization reactions of chalcone and stilbene synthases Biochem J 350 229 235 10926848
Thompson JD Gibson TJ Plewniak F Jeanmougin F Higgins DG 1997 The ClustalX windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res 25 4876 4882 9396791
Page RDM 1996 TREEVIEW: An application to display phylogenetic trees on personal computers Comput Appl Biosci 12 357 358 8902363
Hillis D Bull J 1993 An empirical test of bootstrapping as a method for assessing confidence in phylogenetic trees Syst Biol 42 182 192
Alonso JM Stepanova AN Leisse TJ Kim CJ Chen H 2003 Genome-wide insertional mutagenesis of Arabidopsis thaliana
Science 301 653 657 12893945
Rosso MG Li Y Strizhov N Reiss B Dekker K 2003 An Arabidopsis thaliana T-DNA mutagenized population (GABI-Kat) for flanking sequence tag-based reverse genetics Plant Mol Biol 53 247 259 14756321
Balzergue S Dubreucq B Chauvin S Le-Clainche I Le Boulaire F 2001 Improved PCR-walking for large-scale isolation of plant t-DNA borders BioTechniques 30 496 504 11252785
Boyes DC Zayed AM Ascenzi R McCaskill AJ Hoffman NE 2001 Growth stage–based phenotypic analysis of Arabidopsis Plant Cell 13 1499 1510 11449047
Steinhauser D Usadel J Luedemann A Kopka J 2004 CSB.DB: A comprehensive systems-biology database Bioinformatics 20 3647 3651 15247097
Mirkin B 1996 Nonconvex optimisation and its application. Mathematical classification and clustering, Volume 3 Norwell (Massachusetts) Kluwer Academic Publishers 428 p.
Thioulouse J Dufour AB Chessel D 2004 ade4: Analysis of environmental data: Exploratory and Euclidean methods in environmental sciences. R package version 1.3–3 Available: http://pbil.univ-lyon1.fr/ADE-4 . Accessed 26 August 2005.
R Development Core Team 2005 R: A language and environment for statistical computing Vienna The R Foundation for Statistical Computing
Felsenstein J 2004 PHYLIP (Phylogeny Inference Package) version 3.6 Seattle Department of Genome Sciences, University of Washington Available: http://evolution.genetics.washington.edu/phylip.html . Accessed 26 August 2005.
An YQ McDowell JM Huang S McKinney EC Chamliss S 1996 Strong, constitutive expression of the Arabidopsis ACT2/ACT8 actin subclass in vegetative tissues Plant J 10 107 121 8758981
Baringhaus L Franz C 2004 On a new multivariate two-sample test J Multivariate Anal 88 190 206
Sokal RR Rohlf FJ 1995 Biometry: The principles and practice of statistics in biological research New York W.H. Freeman and Co 887 p.
Efron B Tibshirani RJ 1993 An introduction to the bootstrap New York Chapman and Hall 436 p.
Schoof H Ernst R Nazarov V Pfeifer L Mewes HW 2004 MIPS Arabidopsis thaliana Database (MAtDB): An integrated biological knowledge resource for plant genomics Nucleic Acid Res 32 D373 D376 14681437
|
16193095
|
PMC1236819
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Sep 23; 1(4):e40
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010040
|
oa_comm
|
==== Front
BMC BiochemBMC Biochemistry1471-2091BioMed Central London 1471-2091-6-151611783110.1186/1471-2091-6-15Research ArticleA two disulfide bridge Kazal domain from Phytophthora exhibits stable inhibitory activity against serine proteases of the subtilisin family Tian Miaoying [email protected] Sophien [email protected] Department of Plant Pathology, The Ohio State University, Ohio Agricultural Research and Development Center, Wooster, OH 44691, USA2005 23 8 2005 6 15 15 8 4 2005 23 8 2005 Copyright © 2005 Tian and Kamoun; licensee BioMed Central Ltd.2005Tian and Kamoun; 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
Kazal-like serine protease inhibitors are defined by a conserved sequence motif. A typical Kazal domain contains six cysteine residues leading to three disulfide bonds with a 1–5/2–4/3–6 pattern. Most Kazal domains described so far belong to this class. However, a novel class of Kazal domains with two disulfide bridges resulting from the absence of the third and sixth cysteines have been found in biologically important molecules, such as human LEKTI, a 15-domain inhibitor associated with the severe congenital disease Netherton syndrome. These domains are referred to as atypical Kazal domains. Previously, EPI1, a Kazal-like protease inhibitor from the oomycete plant pathogen Phytophthora infestans, was shown to be a tight-binding inhibitor of subtilisin A. EPI1 also inhibits and interacts with the pathogenesis-related P69B subtilase of the host plant tomato, suggesting a role in virulence. EPI1 is composed of two Kazal domains, the four-cysteine atypical domain EPI1a and the typical domain EPI1b.
Results
In this study, we predicted the inhibition constants of EPI1a and EPI1b to subtilisin A using the additivity-based sequence to reactivity algorithm (Laskowski algorithm). The atypical domain EPI1a, but not the typical domain EPI1b, was predicted to have strong inhibitory activity against subtilisin A. Inhibition assays and coimmunoprecipitation experiments showed that recombinant domain EPI1a exhibited stable inhibitory activity against subilisin A and was solely responsible for inhibition and interaction with tomato P69B subtilase.
Conclusion
The finding that the two disulfide bridge atypical Kazal domain EPI1a is a stable inhibitor indicates that the missing two cysteines and their corresponding disulfide bond are not essential for inhibitor reactivity and stability. This report also suggests that the Laskowski algorithm originally developed and validated with typical Kazal domains might operate accurately for atypical Kazal domains.
==== Body
Background
Proteases play essential roles in biological systems, not only digestion and protein turnover but also a diversity of specific processes [1]. To regulate the activity of proteases and avoid cellular damage, organisms also produce protease inhibitors [1]. So far, 48 distinct families of protease inhibitors have been described, one of which is the Kazal family of serine protease inhibitors (I1 family) [1]. Kazal type inhibitors are widely distributed in animals, apicomplexans and oomycetes. They are thought to play important roles in maintenance of normal cellular and physiological processes of animals [2,3], and pathogenesis of mammalian parasitic apicomplexans and plant pathogenic oomycetes [4-7]. Kazal-like serine protease inhibitors are defined by a conserved motif in their amino acid sequences. Typical Kazal domains contain six cysteine residues forming a 1–5/2–4/3–6 disulfide bond pattern [3,8]. Most Kazal domains described so far belong to this class. However, a novel class of Kazal domains has been described in recent years, in which the third and sixth cysteines are missing resulting in the loss of the 3–6 disulfide bond [3,7,9]. These two disulfide bridge domains are referred to here as atypical Kazal domains.
Atypical Kazal domains were first reported in the human serine proteinase inhibitor LEKTI, a 15-domain inhibitor associated with the severe congenital disease Netherton syndrome [2,9]. Domain 2 and 15 of LEKTI are typical Kazal domains with complete 6 cysteine residues, whereas the remaining 13 domains represent atypical two disulfide bridge Kazal domains [3,9,10]. The functionality of some atypical Kazal domains from LEKTI has been examined. Domain 1 of LEKTI does not inhibit any of the standard proteases [11]. Domain 6 exhibits significant inhibitory activity on trypsin, but this inhibition is only temporary [3,9,11]. A recombinant protein containing four atypical domains of LEKTI (domain 6, 7, 8 and 9) inhibits both trypsin and subtilisin A permanently [10], indicating that atypical Kazal domains can be effective inhibitors. However, it is unclear whether a single atypical domain can be a stable inhibitor. Multi-domain interactions could be responsible for the stable inhibitory activity observed for the recombinant protein [10]. Additional structural and functional studies on atypical Kazal domains are needed to understand the impact of the disulfide bridges on inhibitor activity and stability.
As a result of exhaustive biochemical studies of the third domain of turkey ovomucoid protein performed by the late Michael Laskowski Jr. and collaborators, much is known about the relationship between domain sequence and inhibition specificity in Kazal inhibitor-serine protease interactions. This work culminated in the development of an additivity-based sequence to reactivity algorithm, referred to from here on as the Laskowski algorithm, that predicts the inhibition constants (Ki) between Kazal domains and a set of six serine proteases based solely on the sequence of the inhibitors [12,13]. Structural studies of Kazal domain-protease complexes revealed that there are 12 contact positions (P6, P5, P4, P3, P2, P1, P1', P2', P3', P14', P15' and P18') responsible for interactions between Kazal domains and their cognate serine proteases [12,14-16]. Changes in noncontact residues often do not affect equilibrium constants (Ka, the reciprocal of Ki), whereas changes in contact residues result in significant alterations of Ka [12]. Among the 12 contact residues, P3, the second conserved cysteine residue, and P15', a conserved asparagine, show little variation in naturally occurring Kazal domains, but the remaining ten contact residues are hypervariable [12]. Therefore, the Laskowski algorithm was established based on the residues at the 10 contact positions and allows for the calculation of Ka or Ki of a Kazal domain against a selected set of six serine proteases based on the domain sequence alone [12,13,17]. This algorithm was developed based on 191 variants of turkey ovomucoid third domain (19 amino acid mutants in the ten contact residues plus the wild type) and was validated with a number of typical Kazal domains [12,13,17]. Theoretically, the algorithm should be applicable to atypical Kazal domains since the missing cysteine residues are different from the hypervariable contact residues. However, the accuracy of the algorithm in predicting the reactivity of atypical Kazal domains has not been tested (M. Laskowski, Jr., pers. comm.).
Kazal-like inhibitors are ubiquitous in oomycetes [7], a group of eukaryotic microbes that includes many devastating plant pathogens [18]. A total of 35 putative extracellular proteins with 56 predicted Kazal-like domains were identified from five plant pathogenic oomycete species [7]. Among them, the Phytophthora infestans Kazal inhibitors EPI1 and EPI10 inhibit and interact with the pathogenesis-related P69B subtilisin-like serine protease of the host plant tomato, suggesting a role in virulence [7,19]. Both EPI1 and EPI10 contain an atypical Kazal domain [7,19]. Atypical domains are common in Kazal-like inhibitors of plant pathogenic oomycetes. One-fourth (14/56) of oomycete Kazal domains belong to this type [7]. These domains are distributed in 14 different proteins from three Phytophthora species, P. infestans, P. ramorum and P. sojae, some of which have multiple domains. Remarkably, phylogenetic analysis of the 56 domains revealed that all the atypical domains form a significantly distinct cluster (M. Tian, Z. Liu and S. Kamoun, manuscript in preparation), suggesting that the loss of one disulfide bridge in Phytophthora Kazal-like domains predates speciation. Characterizing the atypical Kazal domains would help to understand the biochemical and biological functions of these inhibitors.
P. infestans EPI1 is an ideal candidate to characterize atypical Kazal domains. EPI1 was identified as a tight-binding inhibitor of subtilisin A and inhibits and interacts with P69B subtilisin-like serine protease [7]. EPI1 is composed of two putative Kazal domains, atypical domain EPI1a and typical domain EPI1b [7]. The predicted 12 contact residues of both domains follow the Kazal consensus, with P3 and P15' conserved cysteines and asparagines, respectively, and the remaining 10 contact residues variable relative to other Kazal domains. In this study, we predicted the inhibition constants of EPI1a and EPI1b to subtilisin A using the Laskowski algorithm [12]. The atypical domain EPI1a, but not the typical domain EPI1b, was predicted to be a strong inhibitor of subtilisin A. A recombinant EPI1a exhibited stable inhibitory activity against subilisin A and appeared solely responsible for inhibition and interaction with tomato P69B subtilase, providing evidence that the missing two cysteine residues and the corresponding disulfide bond might not be essential for inhibitor reactivity and stability. This report also suggests that the additivity-based sequence to reactivity algorithm (Laskowski algorithm) originally developed and validated with typical Kazal domains might operate accurately for atypical domains.
Results
The atypical Kazal domain of EPI1 is predicted to be a functional inhibitor of subtilisin A
The inhibition constants of two Kazal domains of EPI1 against subtilisin A were predicted using the Laskowski algorithm [12] based on the sequence of their 10 hypervariable contact residues (Fig. 1). Interestingly, the atypical Kazal domain EPI1a was predicted to be a strong inhibitor of subtilisin A with a Ki of 4.3 nM, a value that is remarkably similar to the experimentally determined Ki of 2.77 +/- 1.07 nM for the entire EPI1 protein against subtilisin A [7]. In contrast, the typical Kazal domain EPI1b, which contains the complete set of six cysteine residues, may not be functional against subtilisin A since the predicted Ki was high at 50 mM. Therefore, these computational analyses predicted that the atypical domain EPI1a is solely responsible for the inhibition of subtilisin A.
Figure 1 Primary structure alignment of two Kazal domains of EPI1 and the predicted inhibition constants against subtilisin A. The conserved cysteine residues in both domains are shown in bold. The putative P1-P1' sites and the disulfide linkages predicted based on the structure of other Kazal domains are shown. The putative 10 hypervariable contact residues are marked with asterisks. The numbers represent the inhibition constants of two Kazal domains against subtilisin A predicted with the Laskowski algorithm.
Expression and purification of the two Kazal domains of EPI1
To test the protease inhibitory activities of EPI1a and EPI1b, the two Kazal domains of EPI1, and assess the predictions of the Laskowski algorithm, we expressed and purified the two recombinant domains in Escherichia coli as fusion proteins with the FLAG epitope tag at the amino-terminus. The sequences of the recombinant proteins are shown in Fig. 2A. The predicted molecular mass for FLAG-EPI1a (rEPI1a) and FLAG-EPI1b (rEPI1b) was 10181 Da and 8996 Da, respectively. To determine the purity of the purified recombinant proteins, we ran 0.5 μg of purified rEPI1a and rEPI1b on SDS-PAGE gel and stained with silver nitrate. Bands of the expected sizes were observed for both proteins. There was only a single band for rEPI1b, indicating high purity (Fig. 2B). The rEPI1a sample revealed two closely-migrating bands (Fig. 2B). The two bands reacted to the FLAG antibody and are likely to represent rEPI1a with and without the signal peptide OMPA, which is located immediately before the FLAG peptide in the vector pFLAG-ATS and is responsible for secreted expression in E. coli. Similar release of the mature protein was commonly observed with other proteins expressed using pFLAG-ATS (M. Tian and S. Kamoun, unpublished). We also stained the gel loaded with purified rEPI1a protein with Coomassie blue. Compared with the band corresponding to the rEPI1a without OMPA, the slower-migrating band was much weaker (data not shown), indicating the secreted version of rEPI1a was the major component of the purified rEPI1a protein solution. Besides these two bands, no other proteins were detected by silver staining suggesting that the rEPI1a preparation was highly pure.
Figure 2 Heterologous expression of two Kazal domains of EPI1. A, Amino acid sequences of recombinant Kazal domains rEPI1a and rEPI1b. The letters in upper case represent the amino acid sequence of the EPI1 protein. Residues in bold correspond to the native Kazal domains EPI1a and EPI1b as shown in Fig. 1. The letters in lower case represent the vector derived sequence with the underlined ones representing FLAG epitope tag. Numbers indicate the position of amino acid residues starting from the N terminus of EPI1 protein. B, Affinity purified recombinant Kazal domains visualized on SDS-PAGE stained with silver nitrate. The rEPI1a with the signal peptide OMPA is indicated by an asterisk. The numbers on the left represent the size of molecular weight markers.
The atypical Kazal domain EPI1a inhibits the serine protease subtilisin A
We performed inhibition assays of subtilisin A by incubating 0.2 μM of subtilisin A with 0.2 μM of rEPI1a, rEPI1b or buffer control in a volume of 50 μl. The remaining protease activity was measured using the QuantiCleave™ Protease Assay Kit as described in methods. In repeated assays, rEPI1a was found to inhibit about 91% of the measured activity of subtilisin A, whereas rEPI1b did not display any significant inhibition (Fig. 3). These results are consistent with the prediction of the Laskowski algorithm (Fig. 1). It is unlikely that the FLAG tag interfered with the inhibitory activities considering that both the active rEPI1a and inactive rEPI1b carry the FLAG sequence and that these experiments were conducted in parallel.
Figure 3 The atypical Kazal domain EPI1a inhibits subtilisin A. The remaining protease activity of subtilisin A was measured after incubating with rEPI1a, rEPI1b or without protease inhibitors (Std) using the QuantiCleave™ Protease Assay Kit as described in the methods. Activity is expressed as a percentage of total protease activity in the absence of protease inhibitors. The bars correspond to the mean of three independent experiments with three replications for each experiment. The error bars represent the standard errors calculated from the mean of three experiments.
EPI1a inhibits the tomato pathogenesis-related P69B subtilase
EPI1 inhibits and interacts with the pathogenesis-related P69B subtilase of tomato [7]. To test which of the two domains of EPI1 inhibits P69B, we first used agroinfiltration to transiently express P69B fused with the epitope tag HA at the C-terminus in Nicotiana benthamiana leaves. Intercellular fluids were collected from leaves infiltrated with either Agrobacterium tumefaciens containing pCB302-P69B, or A. tumefaciens containing the empty binary vector pCB302-3. 10 μl of intercellular fluids from the two treatments were used in in-gel protease assays. A distinct additional protease band was observed in the P69B-expressing sample but not in the control suggesting that P69B-HA is functional (Fig. 4A). 10 μl of P69B-expressing N. benthamiana intercellular fluids were incubated with 20 pmol of the EPI1 recombinant protein rEPI1 [7], rEPI1a, rEPI1b, or buffer and the remaining protease activity was detected by in-gel protease assay. rEPI1a, containing the atypical Kazal domain, completely inhibited the P69B band similar to rEPI1. rEPI1 and rEPI1a also inhibited the activity of two other extracellular proteases from N. benthamiana (Fig. 4B). The identity of these N. benthamiana proteases is unknown but they could also be subtilisin-like serine proteases, such as homologs of tomato P69. In these experiments, rEPI1b did not exhibit any inhibition towards P69B or other protease bands.
Figure 4 The atypical Kazal domain rEPI1a inhibits P69B subtilase. A, In-gel protease assay of Nicotiana benthamiana intercellular fluids expressing the empty binary vector pCB302-3 (-) or pCB-P69B (+). B, Inhibition assay of P69B by recombinant EPI1 entire protein and the single Kazal domains. P69B-expressing N. benthamiana intercellular fluids were incubated in the presence of rEPI1, rEPI1a, rEPI1b or the absence of protease inhibitors (Buffer) and then the remaining protease activity was analyzed using zymogen in-gel protease assays. The arrow indicates the band location corresponding to the protease activity of P69B.
EPI1a interacts with tomato P69
We previously showed that rEPI1 interacts with P69B [7]. Here, we tested whether the atypical Kazal domain EPI1a interacts with P69B subtilase by coimmunoprecipitation. Coimmunoprecipitation was performed on BTH-induced tomato intercellular fluids incubated with rEPI1a, rEPI1b or buffer control using FLAG antibody covalently linked agarose beads. Western blots were hybridized sequentially with P69 and FLAG antisera and revealed that P69 subtilases co-precipitated with rEPI1a (Fig. 5). This indicates that rEPI1a interacts with P69 subtilases. Since the P69 family of subtilases have at least six homologs (P69A-P69F) [20,21] and the peptide used to generate the P69 antisera is conserved among several homologs [7], we cannot conclude that P69B subtilase is the only protein that was pulled down with rEPI1a. In contrast to rEPI1a, rEPI1b could not be detected after coimmunoprecipitation with BTH-induced tomato intercellular fluids (Fig. 5). However, rEPI1b was detected in control coimmunoprecipitations with the extraction buffer of tomato intercellular fluids (data not shown), suggesting that this protein is not stable in BTH-induced tomato intercellular fluids.
Figure 5 Coimmunoprecipitation of the recombinant Kazal domains and P69 subtilases using FLAG antisera. Eluates from coimmunoprecipitation of rEPI1a, rEPI1b or buffter with proteins in BTH-treated tomato intercellular fluids were run on SDS-PAGE gel followed by sequential immunobloting with P69 (α-P69) and FLAG (α-FLAG) antisera at a dilution of 1:3000.
EPI1a exhibits stable inhibitory activity
To determine whether EPI1a is a temporary or stable inhibitor of subtilisin, we performed stability analyses by incubating subtilisin A with or without rEPI1a for increasing periods of time and measuring the remaining protease activity. To determine the optimal concentration of EPI1a for the stability analyses, we first performed inhibition assays with varying concentrations of EPI1a (Fig. 6A). The concentration of 0.15 μM of EPI1a resulted in inhibition levels of about 80% of the measured protease activity and was selected for the stability analysis. This concentration is within the linear part of the curve (Fig. 6A), suggesting that hydrolysis of rEPI1a could be easily detected as a decrease in inhibitory activity. The inhibitory activity of rEPI1a did not show any decrease over 3 hours of incubation with subtilisin A (Fig. 6B), indicating that EPI1a is a stable inhibitor of subtilisin A. These results are in sharp contrast with those reported for domain LD-6 of LEKTI, which lost 50% of inhibitory activity after 1 hour of incubation with trypsin and lost all inhibitory effect after 3–4 hours [3].
Figure 6 The atypical Kazal domain rEPI1a exhibits stable inhibitory activity against subtilisin A. A, Protease activity of subtilisin A (0.2 μM) in the presence of rEPI1a in concentrations ranging from 2 nM to 0.3 μM. Activity is expressed as a percentage of total protease activity in the absence of protease inhibitors. B, Protease activity of subtilisin A (0.2 μM) after preincubation with 0.15 μM of rEPI1a (black column) or without protease inhibitors (gray column) for a period of time ranging from 30 min to 180 min. Activity is expressed as a percentage of total protease activity in the absence of protease inhibitors at each treatment. The bars correspond to the mean of three independent replications of one representative experiment out of three performed. The error bars represent the standard errors calculated from the three replications.
Discussion
Atypical Kazal domains with two disulfide bridges occur in biologically important molecules. The 15-domain human serine protease inhibitor LEKTI that carries 13 atypical Kazal domains is associated with the severe congenital disorder Netherton syndrome [2]. The protease inhibitors EPI1 and EPI10 of P. infestans contain an atypical Kazal domain and have been implicated in virulence of this devastating plant pathogen [7,19]. Although the structure and function of atypical Kazal domains from LEKTI have been studied, the effects of the loss of Cys 3, Cys 6 and the corresponding disulfide bond on inhibitor reactivity and stability was not assessed and there is no evidence showing that a single atypical Kazal domain can be a stable inhibitor by itself [3,10,11]. In this study, we describe that the atypical domain EPI1a of the two-domain EPI1 protein is a stable inhibitor of the subtilisin family of serine proteases. No loss of inhibitory activity was found even after incubating EPI1a with subtilisin A for 3 hours. The loss of Cys 3, Cys 6 and the corresponding disulfide bond does not have major adverse effects on inhibitory activity or stability, indicating that these two cysteine residues might not be essential for the function of Kazal domains. This finding is important for determining the biochemical and biological functions of Kazal inhibitors containing atypical Kazal domain(s). Kazal-like proteins have been reported from animals, apicomplexans, oomycetes, as well as the bacterium Nitrosomonas europaea [7]. The availability of genomic sequence from a diversity of organisms is likely to reveal an increasing number of atypical Kazal inhibitors. For example, so far, a total of 14 atypical Kazal domains have been identified in plant pathogenic oomycetes [7].
The structural mechanism underlying the stability of EPI1a is not clear. The three-dimensional structure of the atypical domain 6 (LD6) of LEKTI was determined. The overall structure of LD6 resembles the three-dimensional fold of typical Kazal-type inhibitors, but the backbone geometry of its canonical loop is not well defined, providing a possible explanation for its temporary inhibitory activity [11]. There are 13 residues between the first cysteine residue (Cys 1) and the second one (Cys 2) in LD6, instead of 6–9 in most typical Kazal domains [11]. The lack of one disulfide bond and the longer sequence stretch between the first two cysteines were proposed to be the factors responsible for the instability of LD6 [11]. Indeed, the longer sequence stretch between the first two cysteines could explain the difference between EPI1a and LD6. There are only 3 residues between Cys 1 and Cys 2 in EPI1a, which is shorter than in most Kazal domains. The longer sequence stretch of LD6 might lead to the abnormal canonical loop and the non-permanent inhibitory activity. Future work to determine the three-dimensional structure of EPI1a and compare it with LD6 of LEKTI and other Kazal domains should help to unravel the structural mechanism underlying the functionality of atypical Kazal domains.
The Laskowski algorithm was developed and validated based on typical Kazal domains [12,13]. The algorithm exploits an exhaustive analysis of all amino acid variants in the ten hypervariable contact residues of turkey ovomucoid third domain, a typical Kazal domain. The accuracy of the algorithm in predicting the reactivity of atypical Kazal domains has not been evaluated (M. Laskowski Jr., pers. comm.). Here we found that the algorithm correctly predicted which of the two EPI domains is likely to inhibit subtilisins. Our experimental data showed that the atypical Kazal domain EPI1a inhibited subtilisin A, and inhibited and interacted with P69 subtilase similar to the entire EPI1 protein [7]. Also, using the algorithm, the atypical Kazal domain EPI1a was predicted to be a strong inhibitor of subtilisin A with a predicted Ki of 4.3 nM, which was in very good agreement with the experimentally determined Ki of 2.77 +/- 1.07nM [7]. As expected from the predicted Ki of 50 mM, the typical EPI1b domain was not an effective inhibitor of subtilisin A. In summary, it appears that the Laskowski algorithm operates accurately for atypical Kazal domains such as EPI1a. Perhaps, this is expected since the Cys 3 and Cys 6 residues of typical Kazal domains are not contact positions. Nonetheless, our observations and the concordance between predicted and experimental data suggest that gross structural changes that could result from the loss of one disulfide bridge in atypical Kazal domains may not affect the specificity of the interactions between Kazal domains and their cognate serine proteases.
Atypical Kazal domains are ubiquitous in serine protease inhibitors of plant pathogenic oomycetes. Fourteen of a total of 56 Kazal-like domains identified in five plant pathogenic oomycetes have only four cysteines. Two of these Kazal-like inhibitors EPI1 and EPI10 of P. infestans target the defense-related protease P69B of the host plant tomato. The first atypical Kazal domain of EPI1 appears to be solely functional in inhibiting and interacting with P69B. In the three-domain EPI10, the second domain is also an atypical domain that was predicted to be functional against subtilisin based on the Laskowski algorithm [19]. These findings raise some interesting questions. What are the biochemical and biological implications of the loss of the disulfide bridge? Are there any evolutionary advantages of the two-disulfide bridge Kazal domain over the three-disulfide bridge domain in counteracting and co-evolving with host proteases? Additional functional and structural studies are needed to address these questions.
Conclusion
In this study, the functionality of a two disulfide bridge atypical Kazal domain EPI1a from Phytophthora was characterized. EPI1a was predicted to be a strong inhibitor of subtilisin A using the additivity-based sequence to reactivity algorithm (Laskowski algorithm). Inhibition assays and coimmunoprecipitation experiments showed that recombinant domain EPI1a exhibited stable inhibitory activity against subilisin A and was solely responsible for inhibition and interaction with tomato P69B subtilase, providing evidence that the missing two cysteines and their corresponding disulfide bond are not essential for inhibitor reactivity and stability. This report also suggests that the Laskowski algorithm originally developed and validated with typical Kazal domains might operate accurately for atypical Kazal domains.
Methods
Prediction of inhibition constants
The putative ten hypervariable contact residues of EPI1a and EPI1b were identified based on similarity to canonical animal Kazal domains [14-16] and are shown in Fig. 1. Predicted inhibition constants for the EPI1 domains against subtilisin A (Carlsberg) were generated by Drs. M. A. Qasim and M. Laskowski Jr., Purdue University, with the additivity-based sequence to reactivity algorithm (Laskowski algorithm) described by Lu et al. [12].
Plant growth and BTH treatment
Tomato (Lycopersicon esculentum) cultivar Ohio 7814 and N. benthamiana plants were grown in pots at 25°C, 60% humidity, under 16 hour-light/8 hour-dark cycle. We used the salicylic acid analog benzo-(1,2,3)-thiadiazole-7-carbothioic acid S-methyl ester (BTH) to induce PR proteins. BTH treatment of tomato plants followed the exact same procedure described previously [7].
Bacterial strains and plasmids
E. coli XL1-Blue and A. tumefaciens GV3101 were used in this study and were routinely grown in Luria-Bertani (LB) media [22] at 37°C and 28°C, respectively. Plasmids pFLAG-EPI1a and pFLAG-EPI1b for protein expression were constructed by cloning the PCR amplified DNA fragments corresponding to the coding sequence of Kazal domains EPI1a and EPI1b together with some flanking sequence into EcoRI and KpnI sites of pFLAG-ATS (Sigma, St. Louis, MO), a vector that allows secreted expression in E. coli. The primers used for amplification of epi1a are epi1a-F1(5'-gcggaattcTCAAAGCCCGCAAGTCATCAG-3') and epi1a-R1(5'-gcgggtaccTTACTTGCTGGGAGGCTGCTCGCCAG-3'). The primers used for amplification of epi1b are epi1b-F1(5'-gcggaattcCACCGGTAGCTCCACTGGCGAGCAGC-3') and epi1b-R1(5'-gcgggtaccTTATCCCTCCTGCGGTGTC-3'). The introduced EcoRI and KpnI restriction sites for cloning are underlined. The letters in upper case represent gene specific sequence. The detailed sequence information for the expressed fusion proteins FLAG-EPI1a (rEPI1a) and FLAG-EPI1b (rEPI1b) is shown in Fig. 2A. Plasmid pCB-P69B is a construct with the open reading frame of the P69B gene [GenBank: Y17276] fused with the HA tag (YPYDVPDYA) at the C-terminus cloned into the binary vector pCB302-3 [23] and was described elsewhere [19].
SDS-PAGE and Western blot analyses
Proteins were subjected to 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as previously described [22]. Following electrophoresis, gels were stained with silver nitrate following the method of Merril et al. [24] or with Coomassie Brilliant Blue [22], or the proteins were transferred to supported nitrocellulose membranes (BioRad Laboratories, Hercules, CA) using a Mini Trans-Blot apparatus (BioRad Laboratories, Hercules, CA). Detection of antigen-antibody complexes was carried out with a Western blot alkaline phosphatase kit (BioRad Laboratories, Hercules, CA). Antisera to P69 subtilases were raised against a peptide specific for the tomato P69 family [7]. Monoclonal anti-FLAG M2 antibody was purchased from Sigma (St. Louis, MO).
Expression and purification of rEPI1a and rEPI1b
Expression and purification of rEPI1a and rEPI1b was conducted as described previously for other pFLAG-ATS derived constructs [7,25]. Protein concentrations were determined using the BioRad protein assay (BioRad Laboratories, Hercules, CA). To determine the purity, 0.5 μg of the purified protein was run on a SDS-PAGE gel followed by staining with silver nitrate.
Transient expression of P69B subtilase in planta
Transient expression of P69B-HA in planta was performed according to the agroinfiltration method described previously [26]. A. tumefaciens strains carrying plasmids pCB-P69B, empty vector pCB302-3 [23], and pCB301-P19 (J. Win and S. Kamoun, unpublished) were used. pCB301-P19 is a construct expressing the P19 protein of tomato bushy stunt virus (TBSV), a suppressor of post-transcriptional gene silencing in N. benthamiana that significantly enhances in planta transient expression [27]. Overnight agrobacteria cultures were harvested by centrifugation at 2000 g for 20 min, and resuspended in 10 mM MgCl2, 10 mM MES (pH 5.6) and 150 μM acetosyringone. Resuspended agrobacteria cultures of pCB-P69B or pCB302-3 with an optical density (OD600) of 1.0 were mixed with equal volumes of a culture of pCB301-P19 with an optical density (OD600) of 2.0. The mixtures were kept at room temperature for 3 hours and then infiltrated into leaves of 6-week-old N. benthamiana plants. Intercellular fluids from infiltrated leaves were isolated 5 days after infiltration.
Isolation of intercellular fluids
Intercellular fluids were prepared from tomato and N. bethamiana leaves according to the method of de Wit and Spikman [28]. For tomato leaves, a 0.24 M sorbitol solution was used as extraction buffer. For leaves from N. benthamiana, a solution of 300 mM NaCl, 50 mM NaPO4 pH 7 [26] was used as extraction buffer. The intercellular fluids were filter sterilized (0.45 μM), and were used immediately or stored at -20°C.
In-gel protease assays
In-gel protease assays were performed with 10% SDS-polyacrylamide gel containing 0.1% (w/v) gelatin (BioRad Laboratories, Hercules, CA) using BIO-RAD's zymogram buffer system as described earlier [7].
Inhibition assays of subtilisin A by EPI1 Kazal domains
Inhibition assays of subtilisin A by EPI1 Kazal domains were performed using colorimetric QuantiCleave™ Protease Assay Kit (Pierce, Rockford, IL). 0.2 μM of subtilisin A (Carlsberg) (Sigma, St. Luis, MO) was preincubated with different amount of purified EPI1 Kazal domains, in a volume of 50 μl buffer for 30 min at 25°C, and then the remaining protease activity was measured following the procedures as described previously [7]. Analysis of the stable inhibitory activity of rEPI1a against subtilisin A was performed by incubating 0.2 μM of subtilisin A with 0.15 μM of rEPI1a in 50 μl buffer (50 mM Tris, pH 8.0) for a time period of 0–180 min at 25°C and then measuring residue enzyme activity.
Coimmunoprecipitation
Coimmunoprecipitation of Kazal domains rEPI1a and rEPI1b with BTH-treated tomato intercellular fluids was performed using the FLAG-tagged protein immunoprecipitation kit (Sigma, St. Luis, MO) as described previously [7]. 100 pmol of purified rEPI1a or rEPI1b were preincubated with 300 μl of tomato intercellular fluids for 30 min at 25°C. 40 μl of anti-FLAG M2 resin was added and incubated at 4°C for 2 h with gentle shaking. The precipitated protein complexes were eluted in 60 μl of FLAG peptide solution (150 ng/μl) and were analyzed by SDS-PAGE and Western blot analyses.
Authors' contributions
MT, designing and performance of wet lab experiments, writing of manuscript. SK, supervision of experimental work, writing of manuscript.
Acknowledgements
We are grateful to Dr. Michael Laskowski Jr. and Dr. M. A. Qasim from Purdue University for predicting the inhibition constants and sharing their expert knowledge on Kazal inhibitors. We also thank Diane Kinney for technical assistance and three anonymous reviewers for useful suggestions. This work was supported by USDA-NRI project OHO00963-SS. Salaries and research support were provided by State and Federal Funds appropriated to the Ohio Agricultural Research and Development Center, the Ohio State University.
==== Refs
Rawlings ND Tolle DP Barrett AJ Evolutionary families of peptidase inhibitors Biochem J 2004 378 705 716 14705960 10.1042/BJ20031825
Magert HJ Kreutzmann P Standker L Walden M Drogemuller K Forssmann WG LEKTI: a multidomain serine proteinase inhibitor with pathophysiological relevance Int J Biochem Cell Biol 2002 34 573 576 11943586 10.1016/S1357-2725(01)00179-0
Kreutzmann P Schulz A Standker L Forssmann WG Magert HJ Recombinant production, purification and biochemical characterization of domain 6 of LEKTI: a temporary Kazal-type-related serine proteinase inhibitor J Chromatogr B Analyt Technol Biomed Life Sci 2004 803 75 81 15026000
Pszenny V Ledesma BE Matrajt M Duschak VG Bontempi EJ Dubremetz JF Angel SO Subcellular localization and post-secretory targeting of TgPI, a serine proteinase inhibitor from Toxoplasma gondii Mol Biochem Parasitol 2002 121 283 286 12034464 10.1016/S0166-6851(02)00045-2
Pszenny V Angel SO Duschak VG Paulino M Ledesma B Yabo MI Guarnera E Ruiz AM Bontempi EJ Molecular cloning, sequencing and expression of a serine proteinase inhibitor gene from Toxoplasma gondii Mol Biochem Parasitol 2000 107 241 249 10779600 10.1016/S0166-6851(00)00202-4
Morris MT Cheng WC Zhou XW Brydges SD Carruthers VB Neospora caninum expresses an unusual single-domain Kazal protease inhibitor that is discharged into the parasitophorous vacuole Int J Parasitol 2004 34 693 701 15111091 10.1016/j.ijpara.2004.01.004
Tian M Huitema E Da Cunha L Torto-Alalibo T Kamoun S A Kazal-like extracellular serine protease inhibitor from Phytophthora infestans targets the tomato pathogenesis-related protease P69B J Biol Chem 2004 279 26370 26377 15096512 10.1074/jbc.M400941200
Laskowski MJ Kato I Protein inhibitors of proteinases Annu Rev Biochem 1980 49 593 626 6996568 10.1146/annurev.bi.49.070180.003113
Magert HJ Standker L Kreutzmann P Zucht HD Reinecke M Sommerhoff CP Fritz H Forssmann WG LEKTI, a novel 15-domain type of human serine proteinase inhibitor J Biol Chem 1999 274 21499 21502 10419450 10.1074/jbc.274.31.21499
Jayakumar A Kang Y Mitsudo K Henderson Y Frederick MJ Wang M El-Naggar AK Marx UC Briggs K Clayman GL Expression of LEKTI domains 6-9' in the baculovirus expression system: recombinant LEKTI domains 6-9' inhibit trypsin and subtilisin A Protein Expr Purif 2004 35 93 101 15039071 10.1016/j.pep.2003.12.004
Lauber T Schulz A Schweimer K Adermann K Marx UC Homologous proteins with different folds: the three-dimensional structures of domains 1 and 6 of the multiple Kazal-type inhibitor LEKTI J Mol Biol 2003 328 205 219 12684009 10.1016/S0022-2836(03)00245-6
Lu SM Lu W Qasim MA Anderson S Apostol I Ardelt W Bigler T Chiang YW Cook J James MN Kato I Kelly C Kohr W Komiyama T Lin TY Ogawa M Otlewski J Park SJ Qasim S Ranjbar M Tashiro M Warne N Whatley H Wieczorek A Wieczorek M Wilusz T Wynn R Zhang W Laskowski MJ Predicting the reactivity of proteins from their sequence alone: Kazal family of protein inhibitors of serine proteinases Proc Natl Acad Sci U S A 2001 98 1410 1415 11171964 10.1073/pnas.031581398
Laskowski MJ Qasim MA Yi Z Additivity-based prediction of equilibrium constants for some protein-protein associations Curr Opin Struct Biol 2003 13 130 139 12581670 10.1016/S0959-440X(03)00013-7
Read RJ Fujinaga M Sielecki AR James MN Structure of the complex of Streptomyces griseus protease B and the third domain of the turkey ovomucoid inhibitor at 1.8-A resolution Biochemistry 1983 22 4420 4433 6414511 10.1021/bi00288a012
Lu W Apostol I Qasim MA Warne N Wynn R Zhang WL Anderson S Chiang YW Ogin E Rothberg I Ryan K Laskowski MJ Binding of amino acid side-chains to S1 cavities of serine proteinases J Mol Biol 1997 266 441 461 9047374 10.1006/jmbi.1996.0781
Laskowski MJ Kato I Ardelt W Cook J Denton A Empie MW Kohr WJ Park SJ Parks K Schatzley BL Ovomucoid third domains from 100 avian species: isolation, sequences, and hypervariability of enzyme-inhibitor contact residues Biochemistry 1987 26 202 221 3828298 10.1021/bi00375a028
Qasim MA Lu W Lu SM Ranjbar M Yi Z Chiang YW Ryan K Anderson S Zhang W Qasim S Laskowski MJ Testing of the additivity-based protein sequence to reactivity algorithm Biochemistry 2003 42 6460 6466 12767228 10.1021/bi027186u
Kamoun S Molecular genetics of pathogenic oomycetes Eukaryotic Cell 2003 2 191 199 12684368 10.1128/EC.2.2.191-199.2003
Tian M Benedetti B Kamoun S A second Kazal-like protease inhibitor from Phytophthora infestans inhibits and interacts with the apoplastic pathogenesis-related protease P69B of tomato Plant Physiology 2005 138 1785 1793 15980196 10.1104/pp.105.061226
Jorda L Coego A Conejero V Vera P A genomic cluster containing four differentially regulated subtilisin-like processing protease genes is in tomato plants J Biol Chem 1999 274 2360 2365 9891003 10.1074/jbc.274.4.2360
Jorda L Conejero V Vera P Characterization of P69E and P69F, two differentially regulated genes encoding new members of the subtilisin-like proteinase family from tomato plants Plant Physiol 2000 122 67 73 10631250 10.1104/pp.122.1.67
Sambrook J Fritsch EF Maniatis T Molecular Cloning: A Laboratory Manual 1989 2nd Cold Spring Harbor, N.Y., Cold Spring Harbor Laboratory
Xiang C Han P Lutziger I Wang K Oliver DJ A mini binary vector series for plant transformation Plant Mol Biol 1999 40 711 717 10480394 10.1023/A:1006201910593
Merril CR Goldman D Sedman SA Ebert MH Ultrasensitive stain for proteins in polyacrylamide gels shows regional variation in cerebrospinal fluid proteins Science 1981 211 1437 1438 6162199
Kamoun S van West P de Jong AJ de Groot K Vleeshouwers V Govers F A gene encoding a protein elicitor of Phytophthora infestans is down-regulated during infection of potato Mol Plant-Microbe Interact 1997 10 13 20 9002268
Kruger J Thomas CM Golstein C Dixon MS Smoker M Tang S Mulder L Jones JD A tomato cysteine protease required for Cf-2-dependent disease resistance and suppression of autonecrosis Science 2002 296 744 747 11976458 10.1126/science.1069288
Voinnet O Rivas S Mestre P Baulcombe D An enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virus Plant J 2003 33 949 956 12609035 10.1046/j.1365-313X.2003.01676.x
de Wit PJGM Spikman G Evidence for the occurence of race and cultivar-specific elicitors of necrosis in intercellular fluids of compatible interactions of Cladosporium fulvum and tomato Physiol Plant Pathol 1982 21 1 11
|
16117831
|
PMC1236909
|
CC BY
|
2021-01-04 16:26:24
|
no
|
BMC Biochem. 2005 Aug 23; 6:15
|
utf-8
|
BMC Biochem
| 2,005 |
10.1186/1471-2091-6-15
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2071612237810.1186/1471-2105-6-207DatabaseA protein domain interaction interface database: InterPare Gong Sungsam [email protected] Changbum [email protected] Hansol [email protected] Junsu [email protected] Insoo [email protected] Jungsul [email protected] Dan M [email protected] Donghoon [email protected] Deok-Soo [email protected] Jong [email protected] National Genome Information Center (NGIC), KRIBB, Daejeon, Korea2 Biomatics Lab, Dept. of BioSystems, KAIST, Daejeon, Korea3 Geometric Computing Lab, Division of Computer Science, KAIST, Daejeon, Korea4 Laboratory of Computational and Cellular Biology, Dept. of BioSystems, KAIST, Daejeon, Korea5 MRC-DUNN, Cambridge, UK6 Hanyang University, Seoul, Korea7 Object Interaction Technologies, Inc., Daejeon, Korea8 BiO center, Daejeon, Korea2005 25 8 2005 6 207 207 3 5 2005 25 8 2005 Copyright © 2005 Gong et al; licensee BioMed Central Ltd.2005Gong 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
Most proteins function by interacting with other molecules. Their interaction interfaces are highly conserved throughout evolution to avoid undesirable interactions that lead to fatal disorders in cells. Rational drug discovery includes computational methods to identify the interaction sites of lead compounds to the target molecules. Identifying and classifying protein interaction interfaces on a large scale can help researchers discover drug targets more efficiently.
Description
We introduce a large-scale protein domain interaction interface database called InterPare . It contains both inter-chain (between chains) interfaces and intra-chain (within chain) interfaces. InterPare uses three methods to detect interfaces: 1) the geometric distance method for checking the distance between atoms that belong to different domains, 2) Accessible Surface Area (ASA), a method for detecting the buried region of a protein that is detached from a solvent when forming multimers or complexes, and 3) the Voronoi diagram, a computational geometry method that uses a mathematical definition of interface regions. InterPare includes visualization tools to display protein interior, surface, and interaction interfaces. It also provides statistics such as the amino acid propensities of queried protein according to its interior, surface, and interface region. The atom coordinates that belong to interface, surface, and interior regions can be downloaded from the website.
Conclusion
InterPare is an open and public database server for protein interaction interface information. It contains the large-scale interface data for proteins whose 3D-structures are known. As of November 2004, there were 10,583 (Geometric distance), 10,431 (ASA), and 11,010 (Voronoi diagram) entries in the Protein Data Bank (PDB) containing interfaces, according to the above three methods. In the case of the geometric distance method, there are 31,620 inter-chain domain-domain interaction interfaces and 12,758 intra-chain domain-domain interfaces.
==== Body
Background
Proteins are the most important class of molecules in a cell. Most proteins function by interacting with other molecules, especially other proteins. The interactions among proteins are highly regulated and tightly conserved throughout evolution, [1,2] mainly because unnecessary or unsatisfactory interaction (misinteraction) triggered by random mutations can lead to molecular dysfunction. Therefore, interaction interface regions are under pressure from natural selection and are more conserved [3] compared to other exposed non-interface regions of proteins. Protein "structural interactomics" to map all the protein domain interactions is becoming increasingly important as more complete genome sequences are made available [4-7]. Now scientists can map the whole human interactome bioinformatically [8], using ever-increasing experimental data coming from methods such as yeast two-hybrid analysis. Consequently, a higher resolution molecular interaction analysis is also becoming more important.
Since the 1970s, there has been much effort to determine the principles of protein-protein recognition. Pioneers in the field of protein-protein interaction, such as Chothia and Janin [9], have studied the physical and chemical properties of protein interaction sites that contribute to the recognition processes. Colman et al. [10,11] focused on electrostatic and shape complementarity of interaction interfaces using EC (Electrostatic Complementarity) and shape correlation index, respectively. Argos [12] studied interfaces between protein subunits or protein domains. He not only investigated the physicochemical properties of protein interfaces, but also tried to understand the geometric features of protein interfaces using a spline function [13,14]. Jones and Thornton [15] introduced a surface patch method to find out the parameters that contribute to the process of protein-protein interaction. Chakrabarti and Janin [16,17] investigated the structure of interface region by dissecting it into core and rim based on different solvent accessibility. They also addressed the chemical properties of each region.
Recently, there has been a new trend in the study of protein interfaces. Several groups have introduced computational geometric and topology methods for the study of protein interfaces. Most importantly, the Voronoi diagram [18,19,23] has been used to study interfaces of protein complexes. As early as 1974, Richards [20,21] first introduced the Voronoi diagram as an application for protein structure study, although not specifically as an interface analysis tool.
Despite all the efforts to unveil the underlying principles of protein-protein interaction for over 30 years, there has not been much progress at the fundamental level since the research by Chothia and Janin [9]. The interface data derived from different approaches are not well maintained or widely shared amongst scientists. Fortunately, with the help of faster X-ray crystallography and NMR in structural biology, there has been an increase in the number of known three-dimensional protein structures. This 3D structure information is a good source of data for the study of protein interfaces.
Here, we introduce a large-scale protein interaction interface database called InterPare ( or ). InterPare presents interfaces between protein domains identified by three methods. First, the interface is detected by calculating the geometric distance between subunits of multidomain proteins or protein complexes in the PDB [22,27]. In the second approach, buried protein regions are identified by calculating the accessible surface area (ASA) when they form a complex or an aggregate with other subunits or domains. These buried regions can be accessible to water when they are in a free subunit or one domain state. Finally, interfaces are defined by a geometric and topological approach using the Voronoi diagram [18,19,23]. InterPare presents protein interfaces defined by the Voronoi diagram. The interface structure of queried proteins, in the context of the whole protein configuration, can be viewed with three different molecular viewers on the results page. They are the Chime [24], Jmol [25], and InterFacer [26]. InterPare also provides the atomic coordinate files for protein surface, interior, and interface for further analysis.
Construction and content
Data sets
Proteins in the PDB [22,27] were used to investigate interacting interfaces of protein domains. For a domain definition, we used the Structural Classification of Proteins (SCOP) [28,29]. As of this writing, InterPare uses SCOP 1.65 which is based on around 20,600 PDB entries. The ASTRAL compendium [30,31] provides 3D coordinate files of domains in SCOP. InterPare contains 10,583, 10,431, and 11,010 PDB entries that have been identified as containing interacting interfaces according to geometric distance, ASA, and the Voronoi diagram methods (see interface identification methods below) respectively. Figure 1 shows the extent of PDB data sets covered by each method and their overlap according to the three methods. Interfaces from 10,109 PDB entries can be commonly identified by these three methods. All the interfaces derived by the geometric distance method (green) can also be detected by the Voronoi method (blue) because the latter covers all the multidomain proteins in SCOP (11,010 PDB entries based on SCOP 1.65) by using a mathematical definition of interfaces. The three interface identification methods are explained in the following section.
Figure 1 PDB data coverage and overlap of three interface identification methods. Numbers in a circle represent the number of PDB entries whose interfaces have been identified according to ASA (green), PSIMAP (red), and Voronoi (blue) methods. In the case of multidomain proteins, 24, 474, and 403 PDB entries are exclusively identified by ASA, PSIMAP, and Voronoi method, respectively. The main reason for the exclusive detection results from the different method of interface identification. Interfaces from 298 PDB entries, uniquely identified by the ASA method, are not domain-domain interfaces. They are interfaces between a domain and DNA (RNA), or between a domain and non-domain region from a different chain. The numbers are based on SCOP 1.65. All the PDB entries that belong to each category are downloadable on the InterPare website.
Interface identification methods
We identified interaction interfaces of protein domains by:
1) Calculating the geometric distance between atoms in different domains (PSIMAP method).
2) Detecting the differences of Accessible Surface Area (ASA) from all the residues in two states: the detached individual subunit state and the multimeric state.
3) Calculating Voronoi diagrams.
1. The geometric distance method checks the distance between atoms in two interacting domains.
Two domains are assumed to interact with each other if there are at least 5 residue pairs whose atomic distance falls within 5 Angstrom distance (5-5 rule), according to the PSIMAP algorithm [32-34]. In this method, domain-domain interaction interfaces are defined as a set of atoms satisfying the threshold of the 5-5 rule by using FAC PSIMAP method [35]. We define an amino acid residue as an interface residue if its atoms are within the threshold 5 Angstrom is a threshold based on Van der Waals radii of interacting atoms and a solvent such as water. The distance threshold (5 Å is a default) can be varied by users on the website. As the threshold gets higher the number of interface residue gets smaller. We used SCOP 1.65 as a domain definition. It contains 54,745 domains from 20,619 PDB Entries (August 2003). InterPare, at the time of this writing, contains 26,999 PDB entries (September 2004). At present, there is a faster algorithm available that uses the convex hull concept [36]. However, the present C program was efficient enough in that it took only 15 hours to complete the calculation for all the entries in the PDB. It is based on a distributed linux cluster system with 22 computing nodes each of which has Intel Xeon 3.0 GHz CPU and 2 GB memory. Current PSIMAP program can be freely downloadable from the PSIMAP website [37].
2. The Accessible Surface Area (ASA) method detects protein regions that are buried and hence excluded from a solvent when forming a multimer or a complex.
If two or more subunits form a protein complex or aggregate, they have to lose a portion of area that was accessible by a solvent (typically water). With the ASA method, we define interface residues as residues that have lost more than 1 Å2 solvent accessible surface area (ASA) upon aggregation or complexation [15,38,39]. It can be formulated as follows.
For all residues () in a SCOP domain and their corresponding residues () in a PDB entry, and can be either an interface residue (Interface(, ) = 1) or a non-interface residue (Interface(, ) = 0) based on the difference of ASA in that residue. The threshold (1 Å2 in our case) can be selected by the user on the InterPare website (from 1 Å2 to 5 Å2). As the threshold gets higher, the number of interface residues gets smaller. An interface region, in a domain, that consists of at least 10 interface residues is acceptable, and those having less than 10 residues are considered as artifacts. InterPare only serves domain interaction interfaces having at least 10 interface residues. We calculated the ASA of protein molecules using a program called NACCESS [40,41], an implementation of the algorithm developed by Lee and Richards [42]. It calculates the absolute ASA and the relevant ASA in terms of total residues, side chains, polar atoms, and non-polar atoms. Relative accessibilities, for each residue in a domain or a protein, can be expressed as the ratio of the surface area of a residue in an intact state to that of a residue in an Ala-X-Ala tri-peptide state [43]. Surface residues are defined as those that have a relative ASA of more than 5% [44]. Interior residues are defined as those that have a relative ASA of less than 5%. This threshold can also be chosen on the InterPare website. The default van der Waals radii of atoms were taken from Chothia [43]. We used water of 1.40 van der Waals radii as a solvent. In Figure 2, a protein domain is shown which is divided into three regions (interface, interior, surface) according to the ASA method.
Figure 2 Protein structure with respect to their geometrical region. (a) Schematic diagram representing the interior, interface, and surface of longitudinal section of a protein domain. (b) An example of a 3D structure (SCOP id: d1a25a_) which corresponds to a schematic diagram (a). It shows the three areas of a domain (red: protein surface, blue: protein interior, filled-in space model: interaction interface). Interface regions are represented as a space-fill model to distinguish them from other regions.
3. The Voronoi diagram, also known as Dirichlet Tessellation, has been widely used in the fields of science and engineering. The Voronoi diagram was first introduced as an application for the study of protein structures by Richards [20,21]. There is a report on defining molecular interfaces by Power Diagram; Voronoi Diagram on a weighted point set [45]. We used the same protocol suggested by Varshney et al. [45], but applied our own polygon filtering method and calculated interfaces only between domains instead of calculating them on protein complexes.
First of all, a three dimensional power-diagram P of the atoms was constructed. Each face of the power-diagram P is defined by two adjacent atoms (Figure 3). Power-diagrams generate polygons which are bounded by edges. An edge, represented as a blue solid line in Figure 3, is defined by two atoms each of which belongs to different domains. The construction of such a power diagram, in an average case, will have a time complexity of O(n) (n is number of atoms in the protein) [46,47] where the number of neighbors for any given atom is bounded by a constant.
Figure 3 Power diagram of two different domains in 2D representation. Light blue circles (atoms) are contained in domain A, and green atoms are in domain B. Dotted lines denote Voronoi edges between two neighboring atoms, and solid lines represent the Voronoi geometrical interface between two domains. Any polygon which is adjacent to at least one Voronoi geometrical interface is called an interface-cell. If a cell is an interface-cell, then we call the atom in the cell an interface-atom. Interface-atoms are slightly darker than non-interface atoms. The InterPare database stores all interface-atom information.
To have polygons only close to the interaction region, marginal polygons need to be filtered out because those are irrelevant to the interacting interfaces. We removed all the marginal polygons by using our two-stage polygon filtering method. At first stage, we removed polygons which do not contain edges defined by interface atoms. Interface atoms are those in the interface residue defined by ASA method (see above). The default van der Waals radii of atoms were taken from Chothia [43]. Polygons are further filtered out if they have one or more vertices which are beyond 5 Angstrom distance from the interface atoms. For each face in P (Figure 3), if two atoms defining a face belong to different domains from each other, we call such a face an interface-face. Let us define interface-cells as cells in the power-diagram P that have at least one interface-face. Let us define interface-atoms to be those atoms whose cells are interface-cells. In the InterPare database, all the interface-atoms between two domain pairs are stored in a PDB-style file format.
Utility
InterPare contains protein surface, interior, and interface information from PDB entries. There are three query interfaces to access the information in InterPare. Queries can be 1) keywords, 2) PDB or SCOP IDs, or 3) protein sequences in FASTA format. In the case of a protein sequence, InterPare provides a structural domain assignment module using PDB-ISL [48] and PSI-BLAST [49,50] to assign homologous domains in SCOP to the queried sequence. All the queries are finally assigned to (a) PDB ID(s). Figure 4a shows the search interface in the case of a PDB ID as a query. Relative ASA (see interface identification methods above), in Figure 4a, is a criterion for the protein interior and surface boundaries. There are two options for the interface definition threshold: one for the geometric distance method, and another for the ASA method (See the interface identification method above for the threshold criteria). Figure 4b shows the results of PDB ID '1a25' as a query by the ASA method. It contains protein surface, interior, and interface information. We implemented Chime [24] and Jmol [25] scripts to let users view protein 3D structures in a pop-up window when the links are clicked, as in Figure 4c and 4d. Protein surfaces and interiors are in red and blue, respectively, and the interface is viewed in space-fill mode to distinguish it from other parts of the protein molecule. To view protein structures, the Chime plug-in and a Java runtime environment with Java 3D 1.3.1+ are required. The InterFacer homepage provides files that are required to view molecules with InterFacer. Atom coordinate files of three different regions are available to download. In addition, 1) the size of the interface and surface area, and 2) amino acid compositions on the surface, interior, and interface regions are provided on the results page.
Figure 4 Screen shots from the results page of the InterPare web site. (a) The query box with several options. 'Relative ASA' is a criterion for discerning the interior region of a protein from its surface region. There are two independent criteria for the interface definition: 'ΔASA' for the ASA method, and 'Distance' for the PSIMAP algorithm. (b) Results of querying the database with '1a25' as a PDB ID. 1a25 is a C2 domain from the protein kinase C beta. It consists of two identical domains (d1a25a_ and d1a25b_). Results by the ASA method are shown here. The atomic coordinates for the protein surface (Shell), interior (Core), and interface (Face) regions are downloadable on this page. (c), (d) The 3D structure of protein '1a25' by Chime and Jmol, respectively. Interior regions are blue and surface regions are red. Interface regions are displayed in space-fill model and other non-interface regions are displayed in wireframe with a backbone tracing mode.
Discussion
The protein interfaceome can be defined as the whole set of protein interaction interfaces found in cells. There can be many methods to define such an interface data set. We use the concept of the hierarchical classification of protein domains from SCOP. We extend the SCOP classification to molecular interfaces. The advantage of this approach is that each interface can be classified in the context of domain evolution. SCOP Superfamily is the level of classification where protein structures are clearly known to be related within the classification group. The protein Family level in SCOP is a more functionally relevant class, where each member of the Family is related and functionally similar. Below Family, there are individual domains. We applied three algorithms to find interfaces associated with SCOP. Any protein domain classification system, such as FSSP [51] and CATH [52], can also be used. The main contribution to structural bioinformatics is that interfaces can be searched and compared (hence InterPare) by computer.
We expect that hierarchically similar clusters in the interfaceome will have highly conserved interfaces to maintain their interaction partners. This can provide a new level of functional prediction capability for the designing of novel molecules that can interface with proteins and hence control protein activities.
Conclusion
InterPare is an open and public database server for protein interaction interface information. It contains large-scale interface data for proteins whose 3D-structures are known. We identified 31,620 inter-chain interfaces and 12,758 intra-chain interfaces. At this moment, there are 10,583, 10,431, and 11,010 PDB entries whose domain interaction interfaces have been identified according to geometric distance, ASA, and Voronoi diagram methods, respectively. These interfaces are based on protein domains which are from the SCOP database. By using SCOP, InterPare is tightly associated with the domain classification hierarchy, making the search and lookup convenient.
Availability and requirements
InterPare is available through . InterPare is jointly maintained by the National Genome Information Center (NGIC) of Korea, Object Interaction Technologies, Inc., Daejeon, Korea, and the BiO Center . It is free to any user.
List of abbreviations used
PSIMAP: Protein Structural Interactome map
PDB: Protein Data Bank
SCOP: Structural Classification Of Protein structure
FSSP: Fold classification based on Structure-Structure alignment of Proteins
Authors' contributions
SSG worked on the ASA part, drafted the manuscript, and managed this project. CBP implemented a program regarding the Voronoi diagram. JSK developed the InterPare webpage. ISJ and DMB identified protein interfaces using the PSIMAP algorithm. HSC implemented molecular viewer named InterFacer. JSL made C version of PSIMAP program. DSK supervised the development of Voronoi diagram method. DHO and JB supervised this project and revised the manuscript. All authors have read and accepted the final manuscript.
Acknowledgements
We thank colleagues at Biomatics Lab in NGIC and KAIST. We also thank all the scientists in the field of protein-protein interaction. This project was supported by Biogreen21 program of RDA, R01-2004-000-10172-0 grant of KOSEF, and M1040701000105N070100100 grant of MOST. JB is supported by a grant from KRIBB Research Initiative Program. We especially thank Maryana Bhak for editing this manuscript.
==== Refs
Bolser DM Park J Biological network evolution hypothesis applied to protein structural interactome Genomics and Informatics 2003 1 7 19
Park J Bolser D Conservation of Protein Interaction Network in Evolution Genome Informatics 2001 12 135 140 11791232
Caffrey DR Somaroo S Hughes JD Mintseris J Huang ES Are protein-protein interfaces more conserved in sequence than the rest of the protein surface? Protein Science 2004 13 190 202 14691234
Kim WK Bolser DM Park JH Large scale co-evolution analysis of Protein Structural Interlogues using the global Protein Structural Interactome Map (PSIMAP) Bioinformatics 2004 20 1138 1150 14764552
Bolser DM Dafas P Harrington R Park J Schroeder M Visualisation and graph-theoretic analysis of a large-scale protein structural interactome BMC Bioinformatics 2003 4 45 14531933
Park D Lee S Bolser D Schroeder M Lappe M Oh D Bhak J Comparative interactomics analysis of protein family interaction networks using PSIMAP (protein structural interactome map) Bioinformatics 2005 21 3234 3240 15914543
Moon HS Bhak J Lee KH Lee D Architecture of basic building blocks in protein and domain structural interaction networks Bioinformatics 2005 21 1479 1486 15613386
Kim HG Park J Han KS Predicting Protein Interactions in Human by Homologous Interactions in Yeast Lecture Notes in Computer Science 2003 2637 159 165
Chothia C Janin J Principles of protein-protein recognition Nature 1975 256 705 708 1153006
McCoy AJ Epa VC Colman PM Electrostatic Complementary at Protein/Protein Interfaces J Mol Biol 1997 268 570 584 9159491
Lawrence MC Colman PM Shape complementarity at protein/protein interfaces J Mol Biol 1993 234 946 950 8263940
Argos P An investigation of protein subunit and domain interfaces Protein Eng 1998 2 101 113 3244692
Harder RL Desmarais RN Interpolation Using Surface Splines J Aircraft 1972 9 189 191
Meinguet J Multivariate Interpolation at Arbitrary Points Made Simple J Appl Math Phys 1979 30 292 304
Jones S Thornton JM Analysis of Protein-protein interaction sites using surface patches J Mol Biol 1997 272 121 132 9299342
Chakrabarti P Janin J Dissecting Protein-Protein Recognition Sites Proteins 2002 47 334 343 11948787
Bahadur RP Chakrabarti P Rodier F Janin J Dissecting Subunit Interfaces in Homodimeric Proteins Proteins 2003 53 708 719 14579361
Ban YEA Edelsbrunner H Rudolph J Interface surfaces for protein-protein complexes Proceedings of the Research in Computational Molecular Biology, San Diego 2004 27 31
Poupon A Voronoi and Voronoi-related tessellations in studies of protein structure and interaction Curr Opin Struct Biol 2004 14 233 241 15093839
Richards FM The interpretation of protein structures: total volume, group volume distributions and packing density J Mol Biol 1974 82 1 14 4818482
Richards FM Area, volumes, packing and protein structures Ann Rev Biophys Bioeng 1977 6 151 176 326146
Protein Data Bank
Kim DS Cho YS Kim DG Kim SS Bhak J Lee SH Euclidean Voronoi Diagrams of 3D Spheres and Applications to Protein Structure Analysis Japan Journal of Industrial and Applied Mathematics 2005 22 251 265
Chime
Jmol
InterFacer
Berman HM Westbrook J Feng Z Gilliland G Bhat TN Weissig H Shindyalov IN Bourne PE The Protein Data Bank Nucleic Acid Res 2000 28 235 242 10592235
Structural Classification Of Proteins
Murzin AG Brenner SE Hubbard T Chothia C SCOP: a structural classification of proteins database for the investigation of sequences and structures J Mol Biol 1995 247 536 540 7723011
ASTRAL
Brenner SE Koehl P Levitt M The ASTRAL compendium for sequence and structure analysis Nucleic Acids Res 2000 28 254 256 10592239
Park J Lappe M Teichmann S Mapping Protein Family Interactions: Intramolecular and Intermolecular Protein Family Interaction Repertoires in the PDB and Yeast J Mol Biol 2001 307 929 938 11273711
Han KS Park BK Kim HG Hong JS Park J HPID: The Human Protein Interaction Database Bioinformatics 2004 20 2466 2470 15117749
Lappe M Park J Niggemann O Holm L Generating protein interaction maps from incomplete data: application to Fold assignment Bioinformatics 2001 149 156 11238070
Gong SS Yoon GS Jang IS Bolser DM Dafas P Schroeder M Choi HS Cho YB Han KS Lee SH Choi HH Lappe M Holm L Kim SS Oh DH Bhak JH PSIbase: a database of Protein Structural Interactome map (PSIMAP) Bioinformatics 2005 21 2541 2543 15749693
Dafas P Bolser DM Gomoluch J Park J Schroeder M Using convex hulls to extract interaction interfaces from known structures Bioinformatics 2004 20 1486 1490 15231539
PSIMAP
Jones S Thornton JM Principles of protein-protein interactions Proc Natl Acad Sci 1996 93 13 20 8552589
Jones S Marin A Thornton JM Protein domain interfaces: characterization and comparison with oligomeric protein interfaces Protein Engineering 2000 13 77 82 10708645
NACCESS
Hubbard SJ Thornton JM NACCESS 1993 Department of Biochemistry and Molecular Biology, University College, London
Lee B Richards FM The Interpretation of Protein Structures: Estimation of Static Accessibility J Mol Biol 1971 55 379 400 5551392
Chothia C The nature of the accessible and buried surfaces in proteins J Mol Biol 1976 105 1 14 994183
Miller S Janin J Lesk AM Chothia C Interior and surface of monomeric proteins J Mol Biol 1987 196 641 656 3681970
Varshney A Brooks F Richardson D Defining, Computing, and Visualizing Molecular Interfaces IEEE Visualization 1995 95 36 43
Halperin D Overmars MH Spheres, Molecules, and Hidden Surface Removal The Proceedings of the 10th Annual ACM Symposium of Computational Geometry 1994 113 122
Dwyer RA Higher-Dimensional Voronoi Diagrams in Linear Expected Time The Proceedings of the 5th Annual ACM Symposium on Computational Geometry 1989 326 333
Teichmann SA Chothia C Church GM Park JH Fast assignment of protein structures to sequences using the intermediate sequence library PDB-ISL Bioinformatics 2000 16 117 124 10842732
BLAST
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
Holm L Sander C Mapping the protein universe Science 1996 273 595 602 8662544
Orengo CA Michie AD Jones S Jones DT Swindells MB Thornton JM CATH – A Hierarchic Classification of Protein Domain Structures Structure 1997 5 1093 1108 9309224
|
16122378
|
PMC1236910
|
CC BY
|
2021-01-04 16:27:26
|
no
|
BMC Bioinformatics. 2005 Aug 25; 6:207
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-207
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2121612487210.1186/1471-2105-6-212Methodology ArticleA new dynamical layout algorithm for complex biochemical reaction networks Wegner Katja [email protected] Ursula [email protected] Bioinformatics and Computational Biochemistry, EML Research, Schloss-Wolfsbrunnenweg 33, D-69118 Heidelberg, Germany2005 26 8 2005 6 212 212 20 10 2004 26 8 2005 Copyright © 2005 Wegner and Kummer; licensee BioMed Central Ltd.2005Wegner and Kummer; 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 study complex biochemical reaction networks in living cells researchers more and more rely on databases and computational methods. In order to facilitate computational approaches, visualisation techniques are highly important. Biochemical reaction networks, e.g. metabolic pathways are often depicted as graphs and these graphs should be drawn dynamically to provide flexibility in the context of different data. Conventional layout algorithms are not sufficient for every kind of pathway in biochemical research. This is mainly due to certain conventions to which biochemists/biologists are used to and which are not in accordance to conventional layout algorithms. A number of approaches has been developed to improve this situation. Some of these are used in the context of biochemical databases and make more or less use of the information in these databases to aid the layout process. However, visualisation becomes also more and more important in modelling and simulation tools which mostly do not offer additional connections to databases. Therefore, layout algorithms used in these tools have to work independently of any databases. In addition, all of the existing algorithms face some limitations with respect to the number of edge crossings when it comes to larger biochemical systems due to the interconnectivity of these. Last but not least, in some cases, biochemical conventions are not met properly.
Results
Out of these reasons we have developed a new algorithm which tackles these problems by reducing the number of edge crossings in complex systems, taking further biological conventions into account to identify and visualise cycles. Furthermore the algorithm is independent from database information in order to be easily adopted in any application. It can also be tested as part of the SimWiz package (free to download for academic users at [1]).
Conclusion
The new algorithm reduces the complexity of pathways, as well as edge crossings and edge length in the resulting graphical representation. It also considers existing and further biological conventions to create a drawing most biochemists are familiar with. A lot of examples can be found on [2].
==== Body
Background
With the development of sophisticated experimental technology scientists are trying to understand the huge cellular biochemical network of living cells in its entirety. The complexity of this ambitious goal requires the additional use of computers to be able to analyse the data resulting from high-throughput experiments. Computational approaches include the usage of modelling and simulation of biochemical processes which offers new insights into the way biochemical reactions interact with each other and new perspectives for drug development.
Modelling and simulating increasingly complex biochemical networks, however leads again to masses of data. In order to facilitate the understanding of the results, sophisticated visualisation techniques are required. Therefore, visualisation techniques for use in bioinformatics and computational biochemistry have attracted more and more attention in the last years.
One common example for such visualisations is the graphical representation of biochemical reaction networks/pathways. A graphical representation offers the advantage that the topology of the network which is tightly linked to its function, is easily depicted. This topology information is lost when a researcher is confronted with just a list of biochemical reactions. Of course, graphical representations exist that are hand-made and static (e.g. in biochemistry books).
Many examples of graphical representations on the computer are also static, e.g. KEGG [3]. Being static offers no flexibility in the level of detail or in the exact information depicted by the respective graphs. To get rid of this problem dynamic visualisation techniques arose that enable the user on the fly to see and change only those pathways that are needed at time of viewing [4].
In this paper we concentrate on a new dynamic layout algorithm for the graphical representation of complex reaction networks. In such a graph consisting of nodes and edges, the nodes of the graph represent the compounds and the edges the reactions between these compounds. The direction of an edge shows the direction of the reaction. If an edge points from n1 to n2, n1 is the substrate and n2 the product of the respective reaction.
Often there is a differentiation between two types of compounds, main and side compounds. Main compounds lie on the backbone of the pathway, e.g. in linear pathways they participate in adjacent reactions [5]. All other compounds in this pathway are considered as side compounds.
Dynamic graph layout algorithms try to visualise the graph in such a way that it is easy to survey. This means that crossing of edges is avoided as much as possible. Nodes and labels have to be placed such that they do not overlap. This task can in principle be performed by standard graph layout algorithms, however, there are reasons why this is not suffcient in the context of bioinformatics. One of these reasons is the very high degree of connectivity in complex biochemical networks and the other reason is that there is a certain way that biochemists are used to seeing these graphs that does not match the way a standard layout algorithm would represent a complex biochemical network (see [6] for a detailed discussion). Therefore, several specific dynamic layout algorithms for metabolic pathways have been developed in the past.
The algorithm by Karp et al. [5,7] uses a divide-and-conquer method (project BioCyc). In the first step the graph is decomposed into subgraphs. These subgraphs are drawn according to their topology (linear → hierarchical graph layout, cyclic → circular graph layout, branched → tree layout). In the second step a hierarchical layout algorithm assembles these subgraphs to a whole graph.
BioPath [6,8] is a dynamic electronic version of the Boehringer Biochemical Pathway map by Michal [9,10]. It uses an improved hierarchical layout algorithm [11]. However, BioPath is currently not available. As part of the PathDB project Mendes et al. (personal communication) developed a PathwayViewer that consists of an improved hierarchical [11] and an individual circular layout algorithm. Additionally, they allow the user to edit the final drawing.
The above approaches also make more or less use of additional information in specific underlying databases, e.g. information about the order of reactions, side compounds etc. This is an advantage when aiming at the best graphical representation of data in the database, e.g. metabolic pathways. However, it also often restricts the use of the layout algorithm to these applications. Simulation and modelling tools that also want to make use of sophisticated visualisation techniques cannot rely on additional data in most cases. A graph layout algorithm used in these tools has to work, e.g. on the basis of the information as presented in a SBML file [12]. SBML files only contain explicit data about the individual reactions present in a specific biochemical network. Furthermore, previous approaches first calculate the coordinates of the main compounds and subsequently place the side compounds separately as labels near the edge representing the reaction in which they take part (one single label for each occurrence).
In contrast Rojdestvenski [13] uses a modified spring-embedding layout algorithm [14] for 3D-representations mainly. The algorithm considers main and side compounds as nodes during the layout process but with different priorities. First only the main compounds are placed. Second the algorithm is started again with the main compounds and side compounds at the same time, but with frozen coordinates for the main compounds. In contrast to the other three projects the side compounds are treated as nodes of the graph and their coordinates are determined with the spring-embedder algorithm. Furthermore, each side compound occurs only once instead of one node for each occurrence in the network. However, this approach leads to many edge crossings in a graph with highly connected side compounds.
In 2001 Becker et al. [15] developed a divide-and-conquer method similar to Karp et al. Unlike Karp et al. they only differentiate between cyclic and hierarchical subgraphs in the graph and decompose the subgraphs with the help of a force-directed algorithm [16,17]. However, this method is only able to handle main compounds.
Nevertheless, we chose this algorithm as basis for our work, since cyclic and hierarchical structures are the two basic topologies in which every complex biochemical network/pathway can be separated. In addition, the Becker et al. algorithm is not linked to a specific database and is therefore easily adjusted to different environments and needs.
All of the above algorithms work reasonably well for small to medium sized networks. However, the complexity of studied biochemical networks is increasing. For large and complex pathways the existing pathway layout algorithms often face a problem with respect to the number of edge crossings. Such complex pathways contain highly connected nodes and cycles that share nodes with other cycles. In addition, biological conventions stress the importance of cycles, even small cycles in general since such structures represent important recycling processes and shortcuts in the system. However, existing algorithms do not take these conventions into account. We tested e.g. the Becker et al. algorithm with seventeen elementary reactions of the Peroxidase-Oxidase reaction (PO-reaction [18]). These elementary reactions interact strongly with each other and two recycling processes are present. The Becker et al. algorithm created a tangle of nodes and edges which is impossible to survey (Figure 1, as comparison Figure 2 shows the same reactions drawn by our new algorithm). In the following sections we show how the new algorithm solves these problems, reduces edge crossings and presents the reaction in a way that corresponds more closely with biological conventions. This is achieved by identifying even small cycles and by splitting nodes to improve the readability of the dynamic drawing.
Figure 1 Visualisation of the PO reaction [18]. This picture is a result of the Becker et al. algorithm. The seventeen reactions of this pathway are hardly recognisable.
Figure 2 Visualisation of the PO reaction [18]. The picture is a result of our new layout algorithm and shows the reactions in a much clearer way compared to Fig. 1.
Results
As the starting point for our algorithm we used the implementation of the Becker et al. algorithm that is based on the Java graph library YFiles [19]. In general, our algorithm differs from the Becker et al. algorithm in the following ways: it is able to join and split nodes and to detect smallest cycles or cycles of arbitrary size instead of just the longest one.
In addition, since the Becker et al. algorithm is not able to handle side compounds, we included this possibility. Similar to Karp et al. [7] the definition of side compounds results from a predefined list with compound names (e.g. ATP, NADP, H2O, etc.) That list is editable by the user. Each compound (side and main) is treated as a node in the graph. However, in contrast to Rojdestvenski [13] we place side and main nodes simultaneously which means that as default side compounds have the same priority as the main compounds. Only during the process of cycle search, main compounds are prioritised. This default is chosen, because many examples show that the differentiation between main and side compounds is helpful at times, but often somewhat arbitrary blurring the biochemical reality. Nevertheless, it is also possible to generate a layout without any side compounds.
The list of reactions comprising the biochemical network can be submitted as an SBML [12] or a simple text file (listing all reactions separated by semicolon). It is visualised by a hyper-graph, which means that each reaction is represented by two connected dummy nodes, one is linked with all substrates and the other one with all products of this reaction.
In the following, we will describe in detail how our algorithm works. First, we will show how hierarchical and cyclic subgraphs are found and second how these subgraphs are reassembled to a whole graph.
Identifying subgraphs
This section describes the first part of the algorithm which identifies the cyclic and hierarchical subgraphs of a given pathway. To find joined cyclic subgraphs, nodes which are part of more than one cycle are split. The pseudo-code in Figure 3 describes the modified Becker et al. recursive method to identify circular and hierarchical subgraphs in a given pathway. The first step is to search for the smallest instead of the longest cycle (Figure 4, line 1) As explained above this procedure is chosen, since otherwise biologically relevant information might get lost, since small cycles often represent important recycling processes or short cuts in a pathway. One example is shown in Figure 5. All graphs represent a part of the Peroxidase-Oxidase reaction (PO reaction [18]). The first picture was crafted with a graphic program by a biochemist. The second one was dynamically generated with the Becker et al. algorithm. In this case, the two cycles of the first picture are not easy to depict, because of the emphasis on the longest possible cycle. However, the two cycles represent the two main recycling processes of enzyme intermediates and are therefore crucial for the reaction mechanism. These cycles are shown in the third picture in Figure 5 which is generated by our new layout algorithm.
Figure 3 Pseudo code of the method findSubgraphs. This method searches for all hierarchical and cyclic subgraphs of a given pathway.
Figure 4 Overview of the layout process.
Figure 5 Visualisation of a part of the PO reaction [18] (6 reactions). The layout of the graph that can be seen in the first picture was crafted with the aid of a graphic program by a biochemist, the second one was dynamically generated by the Becker et al. algorithm and the third one by our new algorithm.
To be identified, the smallest cycle must contain at least three compounds. However, this number is adjustable by the user, since there are of course cases where the cycle representing biologically important information is not the absolutely smallest. Therefore, if the first layout depicting the smallest cycle is not of the desired quality, the user can change it by increasing the number of compounds for the cycle search. Furthermore, the cycle must contain all the dummy nodes of each participating reaction.
In the first round of cycle searching only the main compounds and dummy nodes are used to find the smallest cycle. Then the algorithm looks for connected components. Connected components are sets of nodes which are directly or indirectly connected via edges. Thus, there exits a path between each pair of two nodes in the set. The Becker et al. algorithm looks for strongly connected components instead. In contrast to connected components, strongly connected components consider the edge direction. Thus, the above definition applies, however, the path between each pair of nodes in the respective highly connected components is only valid if the direction of the edges is always the same.
For this reason our algorithm is able to find both cycles where all edges have got the same direction (e.g. in Figure 5 the left cycle (PER3+ - COII - COI) in the first picture) and also cycles where edges have got different directions (e.g in Figure 5 the right cycle (PER3+ - PER2+ - COIII) in the first picture). Finally, the breadth first search (BFS) [20] finds the smallest cycle if any exists.
If no cycle is found in the first call of the findSubgraph method, the whole graph will be drawn hierarchically. Otherwise, if one cycle has been found already and the method does not detect a second one with only main and dummy nodes, nodes representing side compounds are also included into the cycle search.
The algorithm keeps distinguishing between these three cases:
• No cycle found → draw the complete graph with a hierarchical layout algorithm (Figure 3, line 17)
• All nodes of the graph belong to the found cycle → draw the complete graph with a circular layout algorithm (Figure 3, line 3)
• Complex graph → draw the found cycle with a circular layout algorithm and separate the remaining nodes of the pathway into further cyclic and hierarchical subgraphs (Figure 3, line 6)
In the first case the Becker et al. algorithm uses a standard hierarchical layout algorithm. We improved this standard algorithm by separating the placement of the nodes into two steps. First the main and dummy nodes are placed by the standard algorithm. Second the side compounds are split to create as many nodes as occurrences in reactions exist. Hence, every node has got only one edge. These nodes are positioned one layer above or under the other end point of the edge according to the direction (top to down).
In the second case the whole graph consists of one cycle and all nodes are therefore positioned by a standard circular layout algorithm.
In the complex graph case the graph consists of various circular and hierarchical subgraphs. Since in all existing layout algorithms each node is part of exactly one subgraph, these algorithms are not able to find cycles which share nodes. Therefore, we added the possibility to split (Figure 3, line 11) and join nodes (Figure 3, line 18 and Figure 4, line 2). Figure 6 shows the urea cycle and a part of the citrate cycle crafted with a graphical program by a biochemist. The Becker et al. algorithm finds the urea cycle and considers the unshared parts of the citrate cycle as hierarchical subgraph. In contrast to this picture our new algorithm finds two cycles joined at argininosuccinate.
Figure 6 Visualisation of the urea cycle and parts of the citrate cycle. The first picture was made manually, the second one was dynamically generated by the Becker et al. algorithm and the third one by our new layout algorithm. All visualisations look similar but the two cycles in the first and third picture are not well represented in the middle one.
This result is achieved by splitting nodes in found cycles which could also be part of another cycle (Figure 3, line 11). These nodes must represent compounds and must have at least four edges, two edges to nodes in the found cycle and two edges to nodes which are not part of this cycle. Dummy nodes are not allowed to split. Each node n is split into n1 (node with all edges connecting nodes in the found cycle) and n2 (node with all remaining edges), see also Figure 7. In this way, several cycles representing biochemical reaction cycles that share a compound can be found and represented accordingly.
Figure 7 Splitting nodes. This figure shows an example how a node, in this case argininosuccinate is split into two nodes. The adjacency list (Adjx) shows which edge belongs now to which node.
When no more cycles are found, the remaining nodes are regarded as hierarchical. Split nodes are joined and the subgraph is inspected to find connected components before the improved hierarchical layout algorithm places the nodes of this subgraph (Figure 3, line 18). Each connected component is considered as one hierarchical subgraph but components with only one node are saved in an extra set (Figure 3, lines 20–24) and are placed separately.
Building the complete graph
In this section the subgraphs are reassembled to a complete graph by
• Joining split nodes if possible (= joining cycles at one node, Figure 4, line 2).
• Search for further cycles in the found subgraphs (Figure 4, lines 3–7).
• Reassembling of the found subgraphs to a complete graph using a force-directed layout algorithm (Figure 4, line 8).
• Reducing edge crossings between subgraphs (Figure 4, lines 9–18).
These processes are described in detail in the following paragraphs.
For the joining of split nodes present in cycles, the algorithm tries to join main nodes with priority over side nodes. Therefore, cycles with split main nodes will be joined before cycles with split side nodes. Joining two nodes means that two cycles are rotated and moved together at these nodes. One node will be deleted from the graph and all its edges will be shifted to the other one (see an example in Figure 5 (PER3+)). Only two cycles are allowed to be joined at the same node because more than two cycles would cause edge crossings.
For the detection of further cycles, the algorithm searches for subgraphs which are connected by at least two edges with another subgraph. These edges must have got different source and target nodes. If such edges exists, the algorithm will determine the shortest path between the nodes of these two edges in both subgraphs. The found nodes of this path are used to build a new cycle. This new cycle must also correspond to the above explained definition of a valid cycle and is then drawn accordingly. The new cycle and the already existing one will be joined at the common nodes. See an example in Figure 6, the two cycles are joined at argininosuccinate and the four dummy nodes of the two reactions in which argininosuccinate participates.
In contrast to the Becker et al. algorithm the final reassembling step only starts after all subgraphs are found and split nodes are joined. The force-directed method takes the cycle with the maximal number of edges to other subgraphs as central subgraph and all remaining subgraphs and single nodes as input. The force-directed algorithm places all subgraphs around the central cycle to build the complete graph. Finally, to reduce edge crossings, all edges between found subgraphs are checked. The list of these edges is sorted by their length in descending order because typically the longer the edge the higher the number of edge crossings. Starting from the longest edge the number of edge crossings is counted for each edge of this list and nodes are split to reduce the number of crossings. This number of allowed edge crossings can also be changed by the user.
Before splitting a node the algorithm checks whether one node of this edge was the result of another splitting operation and checks whether its splitting partner could take over this edge with at most two edge crossings. If this is not possible, the target node of the examined edge is split and a new node is created that has got only this edge and is placed near to the source node. Nodes with only one edge are moved directly near to the other edge endpoint. To generate a planar graph the number of allowed edge crossings can be set to zero but that also increases the number of split nodes.
In addition, the placement of labels is automatically done by the used layout algorithms of the YFiles package.
Discussion
We have presented our new dynamical layout algorithm for metabolic pathways. One of the main differences to existing algorithms is the emphasis on finding small cycles. This results in a biochemically meaningful representation in many cases, since cycles in biochemical networks often stand for important processes like recycling of intermediates, energy or electron carrier producing or futile cycles. Therefore, biochemists are used to seeing these processes as graphical cycles and our algorithm takes care of this convention. For those cases where the smallest meaningful cycle does not match the default settings, these can be easily adjusted.
Our algorithm is able to handle linear, cyclic and complex metabolic pathways considering main and side compounds. A complex pathway consists of diverse hierarchical and cyclic subgraphs. Nodes are split and joined to improve the detection of these subgraphs and to minimise edge crossings. Finally in many cases the drawing reflects the biological context better than previous approaches, e.g. cycles which share at least one node can be found and represented.
Therefore, our algorithm satisfies the following constraints:
1. Considering further biological conventions: By identifying cycles which share at least one node and by splitting nodes, pathways are drawn more accurately and more similarly to the visualisations biologists are used to (e.g. Figure 6).
2. Unequivocal distinction between substrates and products: Each reaction is represented by two connected dummy nodes which allows an exact distinction between products and substrates in each reaction.
3. Complexity reduction: The complexity of pathways is reduced by splitting higher connected nodes (see also above). Thereby it is possible to untangle nets of many edges to one node, see KEGG [21] pyruvate metabolism (pathway no. 00630, e.g. Glyoxylate).
4. Edge crossing minimisation: The edge crossings are minimised in the second splitting phase when one node of an edge with more than two edge intersections is split.
5. Edge length reduction: Both dummy nodes of one reaction must be part of the same subgraph which minimises the distance between two dummy nodes. As mentioned above, compounds are split and placed near to each other to minimise edge crossing which also reduces the edge length. These methods also fulfil the constraint that compounds of the same reaction should be placed near to each other.
Since our algorithm is based on the Becker et al. algorithm, it calculates similar results for the examples optimally represented in the Becker at al. publication. These examples and several other pathways from the databases KEGG [21], BioCyc [22] and PathDB [23] drawn by the new dynamical layout algorithm can be found at [2]. In addition, since we want to support simulation and modelling tools, we used SBML files describing models of biochemical networks from the SBML model repository ([24]) and the model database ([25]).
The Mendes et al. algorithm (PathDB) can only be used by a general user in the context of the respective database and therefore just with the pathways stored in those. For this reason we restricted ourselves to these pathways to make a comparison.
In the case of the Karp et al. (BioCyc) algorithm which normally is also used in the context of a database, we were able to compare the algorithm in an isolated manner, since it was generously supplied by Karp and coworkers. The isolated algorithm performed well on small to medium sized samples, however, faced some problems w.r.t. edge crossings when considering larger or higher connected pathways (data not shown). In addition to the information on the individual reactions, the isolated algorithm also uses information about the order of the reaction events which is absent from model files, e.g. SBML files. However, this probably could be easily circumvented by a preprocessing step of the respective SBML file if wanted.
The Mendes et al. algorithm and to a lesser extend the Karp et al. algorithm usually use additional information about the considered pathway from their database, e.g. the order of reaction events as pointed out above, to simplify the layout process.
Since such information is not available to a simulation/modelling tool, our algorithm relies solely on a list of reactions of the pathway and optionally a predefined list of side compounds. The existing layout algorithms for metabolic pathways treat the side and main compounds of a pathway differently from our algorithm. They all treat the side compounds as labels, which results in the labels overlapping in complex pathways. Although side compounds are part of the graph in our algorithm the algorithm produces similar results compared to the existing algorithms in those cases where the latter produce good results and solves the overlapping problem in the more complex cases.
The user defined list of side compounds naturally influences the treatment of the nodes during the layout process. By editing this list the user should keep in mind that different side compound lists lead to different drawings of the same pathway, see Figure 8 and 9. Since some parts of the algorithm rely on stochastic methods, the same pathway could be represented with different layouts. For example, the force-directed layout algorithm could cause different assemblings of the subgraphs. In addition the breadth first search of the cycle finding process checks the nodes according to their number of edges in descending order. If there are different nodes with the same number, the order is chosen randomly which means a varying order of nodes could lead to different cycles.
Figure 8 Visualisation of glycolysis. In this picture ATP, Pi and NADH are considered as side compounds which shows the known hierarchical structure of the glycolysis.
Figure 9 Visualisation of glycolysis. In contrast to Figure 6 all compounds are treated as main compounds and the result consists of two cycles which are joined at ATP. This example shows the importance of the choice of the side compound and demonstrates its influence on the topology of a pathway.
To finish, some words about the complexity of the algorithm. The bottleneck is the cycle search. In the worst case, a breadth first tree for every node has to be calculated when no cycle of the given definition exists. In this case, the complexity of this process is N3 (N = number of nodes). However, if a cycle exists the complexity is much lower leading to faster results.
In the near future we want to integrate standard layout algorithms independent from YFiles in order to be able to e.g. change straight lines in cycles into curves and we want to integrate enzymes and regulators as nodes.
Authors' contributions
KW carried out the development and implementation of the algorithm and participated in discussions and writing.
UK initiated the project and participated in discussions and writing.
Acknowledgements
We would like to thank Jürgen Hesser, as well as Ralph Gauges, Jürgen Pahle, Ursula Rost, Isabel Rojas and Sven Sahle for helpful discussions. Our special thanks to Peter Karp and Suzanne Mercer Paley for supporting us by making their algorithm available and answering many questions. The project was funded by the Klaus Tschira Foundation (KTF).
==== Refs
SimWiz download 2005
Supplementary material 2005
Kanehisa M Toward pathway engineering: a new database of genetic and molecular pathways Science and Technology Japan 1996 59 34 38
Brandenburg FJ Gruber B Himsolt M Schreiber F Automatische Visualisierung biochemischer Information Proceedings of the Workshop Molekulare Bioinformatik 1998 GI Jahrestagung 24 38
Karp PD Paley S Hunter L, Searls D, Shavlik J Automated Drawing of Metabolic Pathways Proceedings of the Third International Conference on Bioinformatics and Genome Research 1994 AAAI Press 207 215
Schreiber F High quality visualization of biochemical pathways in BioPath Silico Biology 2002 6
Karp PD Paley SM Altman R, Brutlag D, Karp PD, Lathrop R, Searls D Representations of Metabolic Knowledge: Pathways Proceedings of the Second International Conference on Intelligent Systems for Molecular Biology; Mento Park, CA 1993 AAAI Press 225 238
Brandenburg FJ Forster M Pick A Raitner M Schreiber F Wingender E, Hofestaedt R, Liebich I BioPath – Visualization of Biochemical Pathways Proceedings of the German Conference on Bioinformatics; Braunschweig 2001 11 15
Michal G Biochemical Pathways (Poster) 1993 Boehringer Mannheim GmbH – Biochemica
Michal G Biochemical Pathways 1999 Spektrum Akademischer Verlag
Sugiyama K Tagawa S Toda M Methods for Visual Understanding of Hierarchical System Structures IEEE Transactions on Systems, Man and Cybernetics SMC 1981 11 109 125
Hucka M Finney A Sauro HM Bolouri H Doyle JC Kitano H Arkin AP Bornstein BJ Bray D Cornish-Bowden A Cuellar AA Dronov S Gilles ED Ginkel M Gor V Goryanin II Hedley WJ Hodgman TC Hofmeyr JH Hunter PJ Juty NS Kasberger JL Kremling A Kummer U Le Novere N Loew LM Lucio D Mendes P Minch E Mjolsness ED Nakayama Y Nelson MR Nielsen PF Sakurada T Schaff JC Shapiro BE Shimizu TS Spence HD Stelling J Takahashi K Tomita M Wagner J Wang J The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models Bioinformatics 2003 1 524 531 12611808 10.1093/bioinformatics/btg015
Rojdestvenski I Banissi E, Borner K, Chen C, Clapworthy G, Maple C, Lobben A, Moore C, Roberts J, Ursyn A, Zhang J Objective and subjectiv relations in data visualization. Examples from molecular biological data collections Proceedings of Information Visualization 2003 544 548
Kamada T Kawai S An algorithm for drawing general undirected graphs Information Processing Letters 1989 31 7 15 10.1016/0020-0190(89)90102-6
Becker MY Rojas I A graph layout algorithm for drawing metabolic pathways Bioinformatics 2001 17 461 467 11331241 10.1093/bioinformatics/17.5.461
Quinn NR Breuer MA A force directed component placement procedure for printed circuit boards IEEE Transactions on Circuits Systems 1979 11 109 125
Eades P A heuristic for graph drawing Congressus Numerantium 1984 149 160
Olsen LF Gauges R, van Gend C, Kummer U Simulations of Oscillations of NAD(P)H and Reactive Oxygen Species in Neutrophilic Leukocytes Proceedings of the 2nd Workshop on Computation of Biochemical Pathways and Genetic Networks 2001 89 99
yfiles 2003
Cormen TH Leiserson CE Rivest RL Introduction to Algorithms 1990 London: MIT Press
KEGG 2004
BioCyc 2004
PathDB 2004
SBML 2005
Model database 2005
|
16124872
|
PMC1236911
|
CC BY
|
2021-01-04 16:27:27
|
no
|
BMC Bioinformatics. 2005 Aug 26; 6:212
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-212
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2151613139510.1186/1471-2105-6-215Research ArticleDynamic covariation between gene expression and proteome characteristics Sharabiani Mansour Taghavi Azar [email protected] Markku [email protected] Tommi O [email protected] Mauno [email protected] Institute of Medical Technology, FI-33014 University of Tampere, Finland2 Research Unit, Tampere University Hospital, FI-33520 Tampere, Finland2005 30 8 2005 6 215 215 30 9 2004 30 8 2005 Copyright © 2005 Sharabiani et al; licensee BioMed Central Ltd.2005Sharabiani 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
Cells react to changing intra- and extracellular signals by dynamically modulating complex biochemical networks. Cellular responses to extracellular signals lead to changes in gene and protein expression. Since the majority of genes encode proteins, we investigated possible correlations between protein parameters and gene expression patterns to identify proteome-wide characteristics indicative of trends common to expressed proteins.
Results
Numerous bioinformatics methods were used to filter and merge information regarding gene and protein annotations. A new statistical time point-oriented analysis was developed for the study of dynamic correlations in large time series data. The method was applied to investigate microarray datasets for different cell types, organisms and processes, including human B and T cell stimulation, Drosophila melanogaster life span, and Saccharomyces cerevisiae cell cycle.
Conclusion
We show that the properties of proteins synthesized correlate dynamically with the gene expression profile, indicating that not only is the actual identity and function of expressed proteins important for cellular responses but that several physicochemical and other protein properties correlate with gene expression as well. Gene expression correlates strongly with amino acid composition, composition- and sequence-derived variables, functional, structural, localization and gene ontology parameters. Thus, our results suggest that a dynamic relationship exists between proteome properties and gene expression in many biological systems, and therefore this relationship is fundamental to understanding cellular mechanisms in health and disease.
==== Body
Background
Cells react to changing intra- and extracellular signals by dynamically modulating complex biochemical networks, and cellular responses to extracellular signals lead to changes in gene and protein expression. These processes can be monitored using genomics and proteomics methods. A number of supervised and unsupervised clustering techniques are routinely applied to classify and group genes based on their expression profiles [1]. While these approaches are sufficient for a general grouping of genes, they do not explain why various genes are coexpressed or whether different regulatory mechanisms are involved. Some studies have focused on the properties of coexpressed genes, such as chromosomal location [2-4], regulatory regions and promoters [5,6]. Correlations have also been observed for some of the properties of encoded proteins, such as function and classification of expressed proteins, including those annotated in MIPS [7,8], gene ontologies [9-11], and structural classes [7]. Furthermore, proteins encoded by coexpressed genes are more likely to interact than proteins in general [12,13].
Since the majority of genes encode proteins, we investigated possible correlations between protein-related properties and gene expression patterns to identify proteome-wide features indicative of trends common to expressed proteins. For example, because the cytoplasm, nucleus and extracellular space have different physicochemical properties, such as pH, ionic composition, and protein concentration, the properties of the proteins that are targeted to different cellular compartments are also different. Because there is variation in the specific proteins that comprise the various proteomes, it is intriguing to hypothesize that cellular signaling leads to significant changes in the protein properties of cells. This idea is supported by studies of the relationship between the overall properties of proteins and their amino acid composition, which has been correlated with protein surface properties [14], subcellular localization [15-17], protein structural class [18], and thermal stability [19].
Results and discussion
A number of microarray datasets for several different cell types and organisms were analyzed to study possible transcriptome-proteome correlations. Expression studies have revealed certain correlations between genome-related features and coexpressed genes, including co-localization [2-4] and the conservation of 5' regions containing regulatory sequences [5,6]. Cells respond to changes in intra- or extracellular environment by altering gene expression to produce proteins that are appropriate for the response. Here we applied the Spearman linear correlation to monitor covariations between a number of proteome parameters and gene expression levels along a time series. We observed very significant and dynamic correlations in all the datasets we investigated. We investigated several high-quality datasets for different cell types, treatments, and organisms, including human T cell stimulation [20] and B cell stimulation datasets [8,21], yeast cell cycle data [22], and Drosophila melanogaster life cycle data [23].
T cell receptor (TCR) activation on the surface of T cells is essential for mounting an adaptive immune response against viruses and microbes. The TCR is a multiprotein complex that activates a large number of signaling pathways [24]. Both the CD3 subunit and a co-receptor such as CD28 must be engaged for optimal activation [25]. Microarray analysis at the transcriptome level has revealed changes in the expression of a large number of TCR-related genes [20]. We have now developed a robust method to identify significant correlations between gene expression levels and 114 protein properties over six time points following TCR activation. We found that amino acid composition and several other protein properties covary with gene expression. These results indicate, for the first time, that gene expression profiles and the properties of the encoded proteins have an integral and dynamic relationship and that the protein constituents and overall properties of the proteome are tightly linked and regulated.
Time point-oriented analysis
To study transcriptome-proteome correlations, we introduced time point-oriented analysis. Protein sequence-derived parameters were correlated to gene expression separately at each time point. Time point-oriented clustering (TOC), in which the expression of genes at each time point is clustered separately, was used for other features. Fig. 1 presents the general overview of the strategy for the statistical analysis. For sequence-based parameters, direct correlation with gene expression levels was calculated. For other protein features, genes were organized into clusters separately in each time point with TOC after which correlation of the expression and protein parameters was calculated. Following TOC, the Spearman correlation was used to test linear correlations between the median of gene expression level and the protein variables in the corresponding clusters. Due to the nature and scarceness of functional and structural data, it was necessary to cluster the data prior to conducting the correlation test. Moreover, we were able to compare the correlation coefficients by unifying the sample sizes using a fixed number of clusters (20) for each data type. The resulting correlation matrix was clustered to group the data based on the protein characteristics, and then the results were visualized. See Methods for a detailed description.
Figure 1 Schematic overview of the analysis. Schema for data mining and statistical analysis. Correlation analysis at each time point was carried out between calculated protein parameters and gene expression levels. In the time point-oriented clustering (TOC) at each time point, correlations were analyzed between the medians of expression levels of the clusters and the medians of the proportions of variables for corresponding proteins.
Correlation between transcriptome and proteomes
The 114 variables at six time points in two data sets (CD3 and CD3/CD28 costimulation) yielded 1,368 correlation tests. The huge amount of information was organized according to time points and variables, clustered using the Self Organizing Map (SOM) method and visualized as in Figs. 2 and 3. A total of 177 results (13% of the tests) were statistically significant (with a correlation coefficient either ≥ 0.12 or ≤ (-0.12) for sequence-derived parameters, or ≥ 0.7 or ≤ (-0.7) with TOC). The p value for the correlation study and for the permutation test had to be ≤ 0.05. When considering the large numbers of genes in the datasets, the correlation coefficient value of 0.12 is notable especially since the results were statistically significant based on the p values of both the correlation coefficients and permutation tests of 1,000 rounds. The results indicate clear trends and correlations at subsequent time points as well. It is unlikely that the properties of all the proteins would correlate with gene expression; still the increased/decreased production of certain kinds of proteins is very clear and statistically significant.
Figure 2 Correlation analysis of the dataset for stimulated T cells. Visualization of Spearman's linear correlations between gene expression and protein properties in T cells stimulated with CD3 (left) or costimulated with CD3/CD28 (right). The columns represent time points 1 to 6 corresponding to 1, 2, 6, 12, 24 and 48 hr after stimulation. The 114 variables (rows) consist of 20 amino acid proportions described by three-letter abbreviations, 29 sequence- or amino acid composition-derived parameters (panels to the left), 50 gene ontologies (marked by GO at the beginning), four predicted subcellular localizations (specified by SubLoc), and 11 structural parameters from SCOP (panels to the right). Correlation coefficients are color-coded. Red indicates a positive correlation; i.e., increased production of proteins, which is either attributed to a property (within the categories of GO, SCOP, or SubLoc) or to higher proportions of a property (including amino acids and physicochemical parameters), leading to relative enrichment of that property. Green indicates a negative correlation; i.e., reduced production of proteins, which is either attributed to the property or to higher proportions of the property, leading to relative depletion of that property. The magnitude of the correlation coefficients is represented by color intensity, as indicated at the bottom. The range is from -0.12 to 0.12 for sequence-derived parameters and -0.7 to 0.7 for categorical data used in the TOC approach. The numbers in brackets indicate the genes for which gene ontologies, subcellular localization or structural classes were identified. The correlation coefficients are clustered using the SOM method so as to group features manifesting similar patterns.
Figure 3 The most statistically significant observations for CD3-stimulated (left) and CD3/CD28-costimulated (right) datasets. Only those features with p ≤ 0.05 for both the correlation coefficient and permutation test are shown. Features are organized by SOM clustering. Sequence-derived parameters are on top, with the others below. Color-coding is as in Fig. 2. The numbers in brackets indicate the genes for which gene ontologies, subcellular localization or structural classes were identified. Amino acids composition parameters were calculated for all the proteins. The actual number of proteins used in any individual analysis varied due to different amounts of missing data in different experiments and at different time points.
Of particular interest are the characteristics that indicated trends in more than one data set or at more than one time point. Of the 75 and 102 significant results in the CD3 stimulation and CD3/CD28 costimulation experiments, respectively, 44 (25%) were common to both (Fig. 3). Not a single parameter showed opposite correlation between the two experiments, indicating that CD28 activation mainly strengthens the signaling downstream of CD3 stimulation. Very early and late time points after CD3 or costimulation generally yielded stronger correlations. Some examples of these correlations are shown in Fig. 4.
Figure 4 Examples of correlations between gene expression and proteome parameters. Comparison of the medians of gene expression (●) and medians of the observed/expected ratio of proteins associated with functional and structural variables (□). a, Protein amino acid phosphorylation, time point 3 in the CD3 dataset; b, SCOP classification of α/β proteins, time point 6 in the CD3 dataset; c, Gene ontology for receptor activity, time point 5 in the CD3/CD28-costimulated dataset; and d, Gene ontology for protein biosynthesis, time point 3 in the CD3/CD28-costimulated dataset. The graphs indicate both positive and negative correlations. The clusters are arranged so that the medians of gene expression along the time points are in descending order.
Subcellular localization data were obtained from three individual sources: gene ontology [26], SCOP [27], and prediction. The consistency of the results from the three independent sources reflects the overall high quality of the analysis (Figs. 2 and 3). Membrane proteins showed a negative correlation; i.e., they are relatively underrepresented according to each of the three sources, although the extent of significance varies.
Amino acid frequencies for about half of the residues correlated with the level of gene expression. Lysine dominated the (R+K)/total parameter, correlating positively in both the CD3 and CD3/CD28 costimulation datasets. The aliphatic index, which measures the proportion of aliphatic residues (A, I, L, and V), correlated negatively in the CD3 dataset. The individual aliphatic residues, however, generally gave positive correlations, with the exception of L and V. The frequency of S and T correlated negatively with gene expression in both datasets. The data also suggest that aromatic residues do not generally covary with gene expression. The ratio of polar and large, negative, positive, as well as nonpolar and small residues correlated positively in both datasets. The effect was stronger in the costimulation dataset. Nonpolar residues correlated negatively. Overall hydropathy parameters correlated negatively with expression levels, although this type of correlation varied across time points, due most likely to differences in the parameters contributing to the hydropathy scales. The metabolic cost of amino acid production has been previously evaluated; e.g., for Escherichia coli [28]. However, the number of high-energy phosphate bonds required did not correlate with the results shown in Fig. 3. The most costly residues, W, F and Y, were not underrepresented, and the least costly amino acids, G and A (except for A at two time points in CD3/CD28 data), were not enriched.
Of particular interest are the characteristics that yielded negative correlations at early time points and positive correlations at late time points, and vice versa. The trend is from a negative towards a positive correlation for the ratio of positive amino acids and (R+K)/total parameters in the CD3 dataset, and D in the costimulation dataset. Serine showed the opposite trend in the CD3 dataset. The proportion of neutral residues and oxidoreductase activity correlated negatively at early time points, but only in the CD3 dataset. The ratio of nonpolar and small amino acids correlated negatively at early time points in the CD3 dataset and positively only at late time points in the costimulation dataset. These results clearly indicate a strong correlation between gene expression and the amino acid composition of protein products. In particular, polar and charged residues correlate strongly, and there is a biased occurrence of hydrophobic and aliphatic residues. Interestingly, leucine and isoleucine, which are hydrophobic structural isomers, had opposite correlations.
Protein molecular size and weight are weak indicators, because only MWRES (molecular weight per residue) correlated significantly in the costimulation dataset. Of the volume-related parameters, nonpolar and polar volume correlated negatively, and Polfrac max correlated positively in costimulated cells. Protein flexibility correlated positively in the CD3 dataset at one time point. The most flexible residues, G and A, did not correlate with the transcriptome (except for A at two time points).
SCOP [27] contains information about classification of protein structures at four levels. In addition to membrane and cell surface proteins and peptides, intracellular proteins correlated positively with gene expression in the CD3/CD28 costimulation dataset.
Fifty ontology classes had a large number of entries, 26 of which correlated significantly with expression (Fig. 3). Many of the ontologies were related either to signal transduction or subcellular localization (9 and 11 ontology groups, respectively), with 23 significant findings altogether. Surprisingly, ontologies for immunological or inflammatory responses were not significantly enriched. In fact, the ontology for immune response correlated negatively at time point 5 in the coexpression dataset. Fig. 5 shows the gene expression patterns for some of the largest ontologies.
Figure 5 Gene expression in enriched Gene Ontology classes in the T cell dataset. The left array shows the gene ontology group receptor activity (4872), the middle array shows protein amino acid phosphorylation (6468), and the right array shows transcription factor activity (3700). Genes and proteins were annotated extensively. Signaling activities in the first two ontology classes are indicated by color coding of the block to the left of the expression patterns: protein tyrosine kinase (PTK, red), protein serine/threonine kinase (PSK, blue), dual-specificity kinase (DSK, cyan), receptor (R, black), and protein phosphatase (green). Genes are sorted based on their expression at time point 3. The box "◀" indicates the boundaries for significantly overexpressed, unchanged, and underexpressed genes at time point 3. Only 9% of receptor activity genes are significantly overexpressed, whereas 31% are significantly suppressed. All PSKs are located in the middle (i.e., changes in their gene expression at time point 3 are insignificant). Hypergeometric analysis revealed significant overrepresentation (p < 0.0008) of PSKs as well as significant underrepresentation of PTKs and Rs in this group. PTKs are overrepresented among the underexpressed genes (p < 0.005). When the genes are grouped according to the significance of their expression at time point 6, very significant over- and underrepresentation of PSKs is apparent among the groups of over- and underexpressed genes, respectively. A similar hypergeometric analysis for protein amino acid phosphorylation indicates consistently significant PSK underrepresentation and PTK overrepresentation among underexpressed genes at time point 3 and relative PSK enrichment and R depletion at time point 6. Significant depletion of membrane proteins and those integral to the plasma membrane appears at late time points (Figs. 2 and 3).
Intracellular proteins correlated positively towards the end of the time series in both datasets. Membrane proteins in general and those integral to the plasma membrane were significantly underrepresented at several time points. Based on the prediction data, extracellular proteins correlated negatively with gene expression in the CD3/CD28 dataset.
Our results are somewhat related to previous analyses of the yeast proteomes [7,29], data from which indicate that the abundance of some amino acids and certain overall structural and functional properties correlate with protein abundance. However, no statistically significant differences were observed when studying time points in the diauxic shift data [7]. Some of the yeast results [7,29] do not agree with the T cell data (e.g., the enrichment of amino acids). Lymphocytes function as individual cells, and in this respect are similar to unicellular organisms like yeast; however, lymphocyte function and development are critically linked to the presence and activities of other cells. Of the observations made in yeast, only the enrichment of lysine agrees with our results. Some of the differences are undoubtedly due to the different organisms used in these studies, whereas others may have arisen from differences in experimental goals and methods.
The previous yeast study [29] was aimed at calculating the abundance of protein products; however, we did not attempt such calculations due to the lack of experimental data as well as the problems inherent in correlating protein abundance with that of mRNAs [30,31].
T cell signaling
TCR stimulation activates several signaling pathways [24,32,33], and the detailed annotations of the genes having prominent changes in expression that were generated by our analyses facilitated the identification of these processes and pathways. Receptor activity had a significant negative correlation at time point 5 in the costimulation dataset. There were only a few activated receptors, including IgG receptor, IL-7 receptor, some forms of TNF receptor and G protein-coupled receptors (GPCRs) (Fig. 5). Another significantly affected ontology group was protein amino acid phosphorylation (time point 3 in the CD3 experiment), which was the most characteristic property of the numerous signaling-related observations. There were few changes in signaling molecules in the first hour after stimulation. Several known components of the TCR signaling pathway were overexpressed, including those of the MAP kinase, Ca2+-related signaling, and NF-κB pathways. Subsequently, transcription factors became activated, and a significant positive correlation was observed at time point 3 in CD3-stimulated cells. The enriched transcription factors included early growth response factors 1, 2 and 4, NFAT, NF-κB, Jun, Fos, B cell translocation gene, and interferon regulatory factor. Thus, all the major transcription factor components active in TCR signaling were present (i.e., Fos and Jun forming AP-1, NFAT, and NF-κB). In agreement with these results, transcription factors expressed in lymphocytes have highly regulated expression patterns [21].
We analyzed a number of additional datasets to address whether the associations seen in the T cell data analysis represent sporadic observations that are cell type-specific. Dynamic covariation occurred between gene expression profiles and a number of proteome characteristics describing amino acid proportions and physicochemical properties in all the investigated cell types and organisms, suggesting that our T cell data is representative of cells in general.
Human B cell data analysis
The B cell dataset [8,21] represents genes involved in maturation in anti-immunoglobulin M-stimulated Ramos cells. The development of adaptive immunity and responses to foreign molecules and organisms is based on the highly regulated production of hundreds of proteins. B cell maturation is a multi-step process that requires the ordered expression of a large number of genes. B cell differentiation is activated by non-covalent cross-linking of the B cell receptor (BCR), initiating cellular signaling cascades that ultimately activate nuclear transcription factors. Then, transcriptional activation represses or activates gene expression leading to B cell proliferation, upregulation of surface activation markers, and increased antibody synthesis.
Proline and R/(R+K) were significantly enriched during several consecutive late time points, and neutral and small residues were enriched in the last time point (Fig. 6), whereas I, K, Y; and nonpolar and small residues were underrepresented at all late time points. Among the gene ontologies, only transcription factor activity and signal transduction were significantly positively correlated, both at one time point. None of the protein subcellular localization predictions correlated significantly. Maximum correlations are generally seen at late time points, concurrent with the largest alterations in gene expression. This dataset yielded the least amount of significant findings, which is most likely due to the smaller number of genes in the dataset.
Figure 6 Data analysis for human B cell differentiation. (Left) Correlation coefficients for 1,358 genes involved in maturation in Ramos B cells at 11 sequential time points. (Right) Only the highly significant cases are shown.
S. cerevisiae cell cycle data analysis
Microarrays have been used to study gene expression in S. cerevisiae cultures synchronized by three independent methods: α factor arrest, elutriation, or arrest of a CDC15 temperature-sensitive mutant [22]. Expression of several yeast genes is known to vary periodically during the cell cycle, and the corresponding gene products may be involved in processes occurring only once during the cycle. To investigate these genes, cell cultures must be synchronized so that the cells are simultaneously in the same cell cycle stage. Several approaches are available for synchronization, three of which were used in this study. The first was elutriation whereas other two were cyclin-dependent. In addition, the effects of the G1 cyclin, Cln3p, and the B-type cyclin, Clb2p, were tested. Cyclins are special cell cycle regulators. Note that the CDC28 data was obtained from ref [34].
Several significant correlations are apparent in Fig. 7. Almost all the investigated sequence-based parameters covary with gene expression, but correlations at adjacent time points are rare. At some time points, there are no significant correlations, a recognizable difference compared with the other datasets. This leads to patchy patterns of covariation. If we ignore the time points for which there were few or no correlations, we notice that the different cell culture synchronization treatments yielded rather similar patterns. In the elutriation data, the positively correlated parameters are stronger at the beginning of the time series. Another feature typical for the yeast data is that the correlations are stable in the sense that they seldom change from positive to negative (or vice versa) during a treatment.
Figure 7 Data analysis for the yeast cell cycle. (Top) Correlation coefficients for 590 yeast genes in separate experiments. (Bottom) Only significant correlations are shown.
At the early time points, the correlations are positive (especially in the elutriation treatment) for F, G, I, L, V, W, Y, hydropathy values, as well as for ratios of nonpolar and large, nonpolar, and neutral amino acids. Correlations are mainly negative for E, K, (R+K)/total, rho, and ratios of positive, and polar and small residues. The correlation of V to gene expression is consistent with a previous yeast study [7]; however, since we analyzed dynamic changes, amino acids may appear enriched or underrepresented depending on the time point and the effectors.
Large numbers of correlations occur in the CDC15 and CDC28 experiments, in which the largest fraction of significant MIPS functional classifications was observed in a previous study [35]. This implies that there is consistency among the observations regarding physicochemical properties of proteins and their functional attributes in association with gene expression. Only one gene ontology class, cytoplasm, shows significant correlation – at just one time point. Presumably, the reason for the low abundance of significantly correlated ontologies lies in the smaller number of significantly altered genes compared with the other datasets.
D. melanogaster life cycle data analysis
The D. melanogaster study was performed to follow development and gene expression in a multicellular model organism. In this dataset [23], the transcriptional profiles were investigated throughout the life cycle, from fertilization to aging adults. Samples of both males and females were taken during a complete time course of development of wild type fruit flies up to 30 days of adulthood.
There are dynamic correlation patterns between gene expression and proteome properties during the embryonic period and especially the pupae period (Fig. 8). Correlations are relatively stable in larva as well as during adulthood, especially in males. Continuity and strength of the correlations are unique to the D. melanogaster data, in that once a significant correlation occurs it remains for a prolonged period (i.e., over several consecutive time points).
Figure 8 Data analysis of the D. melanogaster life cycle. (Top) Correlation coefficients of 2,976 genes in wild-type flies examined at 66 time points beginning at fertilization and spanning the embryonic, larval, and pupal periods as well as the first 30 days of adulthood. (Bottom) The most significant cases are shown. Two transition periods are apparent from the correlations. The continuity and strength of the correlations are unique to the D. melanogaster data.
The data contain two major transitions of correlations that do not coincide with developmental stages (Fig. 8). The first transition occurs during the embryo stage, from approximately 11 to 20 h, whereas the other occurs during the pupae period, from 4 to 48 h. There is a somewhat more stable period between 12 to 36 h. Interestingly, covariation patterns are very similar between the transition periods. The observed significant correlations from 15–16 h from the embryo stage until the end of the larval stage are quite similar to the time period from 60 h (pupae) through adulthood (especially in males). On the other hand, the pattern from the beginning of the embryonic stage to 10 h has a covariation pattern that is opposite to the two other conserved time periods. Those parameters having positive correlation at the beginning of the embryonic stage are negatively correlated at late embryo/larvae and late pupae/adult time points, and vice versa. If we associate the covariation patterns between proteome and gene expression during the early embryonic and larval periods to growth and differentiation, respectively, we can infer that pupae have a mixture of both patterns, reflecting a transition from growth to differentiation and then back to growth. Only a few significant correlations are apparent during the transitions, suggesting that the gene expression profile is not in balance (i.e., it changes from one stage to another).
The larval and adulthood data show stable and constant patterns, in which correlations occur between gene expression and a large number of parameters, including a negative correlation for charged and polar residues, their ratios, Polfrac max, rho and flexibility. Positive correlations are seen for aliphatic, aromatic and hydrophobic amino acids, aliphatic index, and hydropathy parameters. Contrary to the observations for T cells, I and L show similar trends. As noted above, the patterns of positive and negative correlations are almost mirror images when comparing the first half of the embryo stage with the two constant regions (i.e., late embryo and larval time points or the late pupae and adult period). Almost all protein properties have statistically significant correlations; the exceptions are N, T and the ratio of neutral and small residues, for which significant correlations occur sporadically.
Several gene ontologies (17 classes) have significant correlations. Six ontologies are related to gene expression/DNA binding. They all have positive correlations during the first half of the embryonic stage and negative correlations thereafter. Nuclear localization follows the same pattern. The correlation motifs for the other ontologies are less clear. Oxidoreductase activity correlates negatively at several time points during the embryonic stage and correlates positively at other time points throughout the experiment. The signal transduction ontology has a similar overall pattern, but only in the pupa and adult females.
Conclusion
Strong correlations between transcriptome and proteome characteristics appear in all the datasets we studied. One could not expect to obtain similar results from the different studies, which differed with respect to both the processes (growth, differentiation, stimulation, and cell cycle) and proteins produced. The strength of the correlations is apparently associated with the intensity of gene expression. Thus, proteins having significantly altered expression do not have a random distribution of physicochemical properties. In particular, the yeast and D. melanogaster life cycle datasets have obvious recurring cycles of enrichment and underrepresentation of different properties. The magnitudes of underrepresentation and enrichment are proportional to gene expression in all experiments. These observations along with our findings related to the functional and structural properties of proteins as well as MIPS classification in earlier studies [7,21] suggest that protein characteristics, including function, structure, subcellular location, and physicochemical properties, are closely associated with gene expression. Furthermore, our method clearly detects dynamic shifts in the gene expression profiles, as exemplified in the D. melanogaster dataset.
Similar, although not identical correlations were obtained when analyzing time series datasets for human T and B cells, yeast, and D. melanogaster, indicating that the dynamic correlations between proteome-related parameters and gene expression likely represent a general paradigm. Since the association is strong and is observed at many levels, cells and organisms, and appears to be a widespread phenomenon, it likely reflects fundamentally important biological processes.
Only a few studies of system-wide transcriptome-proteome correlations have been published, including human heart [36] and platelets [37], mouse liver and kidney [38], and mosquito salivary gland [39]. Only functional features, mainly gene ontologies, were investigated in these articles. Certain ontologies were clearly enriched in all the cases. There was statistically significant colocalization of coexpressed genes in the mouse assay.
Proteome-wide isoelectric points and molecular masses, the two properties used for separation in 2D gels, have been analyzed for 103 organisms, including bacteria, archae and eukaryotes [40]. The comparison of properties of theoretical proteomes for 11 bacteria to the usage of 95 different carbon sources indicated that the ecological niche of bacteria correlates with their proteome parameters [40]. Comparable to these organism-wide, macro-level correlations, we observed several micro-level (time point) correlations in the different datasets of our present study.
There are presumably several reasons for the observed enrichment/depletion of protein properties, which might be related, for example, to cellular processes involved in changes in metabolism and signaling, localization of proteins within cells and compartments, and complexes and interactions formed between proteins. There are indications for the organellar enrichment of proteins having certain types of properties; for example, the need for positive charge in DNA-binding proteins (including histones). In certain disease conditions, the overall amino acid composition of proteins changes in cells, organs, or body fluids [41,42]. Thus, proteome parameters might have diagnostic value and may indicate that health and disease states are linked to both the production and properties of expressed proteins.
Methods
Gene expression data were derived from peripheral T cells subjected to seven treatments: mock (untreated) cells; CD3- or CD28-stimulated cells; CD3 and CD28-costimulated cells; and cells treated with ionomycin along with phorbol 12-myristate 13-acetate (PMA), phytohemagglutinin, or FK506. Data were acquitted at six time points up to 48 hr [20]. We did not investigate the chemical treatment results in detail because they were quantitatively similar to CD3/CD28 co-treatment. Because there were only a few genes that had a significant change in expression (>1.5 fold) in the mock and CD28-stimulated datasets, we focused exclusively on the CD3-stimulated and CD3/CD28-costimulated datasets. In total, 4,359 cDNA elements (of about 18,000 genes and ESTs on the chip) representing 2,926 genes were significantly altered after CD3 stimulation or CD3/CD28 costimulation [20]. The data were taken from .
The B cell dataset for genes involved in maturation in anti-immunoglobulin M-stimulated Ramos cells indicated that close to 1,500 genes had significantly altered expression, at least at one time point [8,21]. These data were from our own experiments. In the D. melanogaster microarray data set, the transcriptional profiles were investigated throughout the life cycle [23]. RNA expression levels of 4,028 genes in wild-type flies were examined at 66 time points. Expression of 3,483 out of 4,028 (86%) changed significantly [P < 0.001; analysis of variance (ANOVA) during the 40-day survey period [23]. The data for genes were obtained from from file Arbeitman.SOMtables.zip. Gene expression information was from .
A total of about 600 genes had altered gene expression patterns in yeast cultures synchronized by four independent methods (α factor arrest, elutriation, cdc28, and arrest of a cdc15 temperature-sensitive mutant) [22]. We also studied the effect of treatment with either the G1 cyclin, Cln3p, or the B-type cyclin, Clb2p. The data were collected from and .
Data mining
We used numerous bioinformatics methods to filter and merge information regarding gene and protein annotations in a number of databases and further calculated and predicted a large number of characteristics for each gene/protein. A dedicated database (Siermala et al., unpublished data) constructed for gene and protein information was extensively used for annotations and sequence identification. To identify the corresponding proteins, we used Locus Link ID numbers of ESTs and genes to retrieve UniProt sequences. FlyBase [43] was used to identify D. melanogaster protein sequences, and NCBI genome sequences were used for yeast.
Sequence-derived variables
Physicochemical parameters and amino acid proportions were directly calculated from protein sequences. Amino acids were further investigated by analyzing the proportions of different groups of residues, namely positive (R, H, K), negative (D, E), neutral (A, N, C, Q, G, I, L, M, F, P, S, T, W, Y, V), polar (D, E, H, K, N, Q, R, S, T, Y), and nonpolar (A, C, F, G, I, L, M, P, V, W). Combinations of characteristics were also investigated, namely neutral and small (A, G, P, S, T), nonpolar and large (F, W, Y), nonpolar and small (I, L, M, V), polar and large (H, K, R), polar and small (D, E, N, Q). In addition, the ratio of Glx (E, Q) and Asx (D, N) to total amino acids, the ratio of R to R plus K, and the ratio of R and K to total amino acids were calculated.
Several parameters for physicochemical features described in the literature were calculated. The aliphatic index is defined as the relative volume occupied by aliphatic side chains (i.e., A, I, L, and V) and was calculated as follows
AI = X(Ala) + 2.9 × X(Val) + 3.9 × (X(Ile) + X(leu)),
where X(Ala), X(Val), X(Ile), and X(Leu) are mole percents for the amino acids [44]. Residues were classified as polar and non-polar as per Fisher's definition [45]. The volumes of each category were calculated by summing the products of the number of each residue [46]. rho is the ratio of polar to non-polar volumes. Nonpolar side chains (NPS), the frequency of nonpolar side chains, was calculated according to Waugh's definition by counting W, N, T, F, P, L, and V residues and expressing the sum as a fraction of the total number of residues [47]. MWRES is the molecular weight per residue. The average hydrophobicity of proteins [46] was calculated from:
where ni is number of residue i, and Hi is the hydrophobicity value of the residue. Four different hydropathy scales [48-51] were used.
POLFRAC_MAX (polar fraction) in an extended chain conformation is calculated as:
where NRES is the total number of residues in the protein and POLFRAC_MAX is in units of e, the electronic charge, SURF is maximal accessible surface area, and POLSURF is polar surface area [52]. Molecular weight was calculated from the sequence as well as the number of amino acids. Isoelectric point was calculated with the EMBOSS program iep. In vivo half-life of proteins was calculated according to [53].
Average flexibility [54,55] for each protein was predicted from the amino acid sequence using a 9-residue sliding window averaging technique with the formula:
where and Bnc is the flexibility parameter of the residue in position i.
For the T cell experiment [20], we identified 1,687 amino acid sequences for 2,926 genes. Sequences were found for 761 proteins in human B cell data and 415 for the yeast dataset. For D. melanogaster [23], 2,976 amino acid sequences for 4,028 genes/ESTs with significant changes in expression were identified and included in the analysis. The actual number of proteins used in any individual analysis varied due to different amounts of missing data in different experiments and at different time points.
Subcellular localization
Subcellular localization of proteins was predicted using SubLoc software [56], which yielded predictions for four categories (cytoplasmic, nuclear, mitochondrial, and extracellular). Transmembrane regions were further predicted using the TMHMM server [57]. To assign each protein to a compartment, we used the SubLoc reliability index and the length of the predicted transmembrane region. If the value of the SubLoc prediction parameter was ≥ 5, the assignment was accepted for the most accurate prediction. In cases where the parameter was between 2 and 4, the length of the transmembrane stretch(es) was taken into account. If the transmembrane region was longer than 36, the protein was predicted to be membrane associated. In cases where the SubLoc prediction value was 1 and the transmembrane region was ≥ 18 residues (the average length of a transmembrane α-helix), the protein was predicted to be membrane spanning. We made predictions in T cell data for 443 proteins, 177 of which were transmembrane, 167 nuclear, 67 extracellular, 7 mitochondrial, and 25 cytoplasmic. Predictions were made for 224 proteins in the B cell data.
Gene ontology
Gene ontology [26] information was extracted from Entrez Gene [58] or from annotated genome data for yeast sequences. In total, we identified the cell component ontology for 2,448 proteins, molecular function for 2,760 proteins, and biological processes for 2,748 proteins in the T cell dataset. Ontologies were available for 1,260 entries in the B cell dataset, for 590 in the yeast dataset, and for 2,293 genes/ESTs in the fruit fly dataset.
Structural description
Hidden Markov Models (HMMs), downloaded from Superfamily [59], were used to search SCOP-derived domains [27] against the protein sequences in T cell data using the program HMMER [60], setting the e-value to <10e-6 as a limit. SCOP information was available for 2,022 entries. The limit of e-values was set so that false positives were unlikely. Moreover, we compared a number of findings with those acquired using the InterProScan sequence searching service [61], which yielded nearly identical results.
Correlation analysis
To monitor dynamic covariations between a number of proteome parameters and gene expression levels along the time series, we applied the Spearman linear correlation test. The schema of the analysis is presented in Fig. 1. The value of the correlation coefficient is dependent on the number of genes/proteins. A value of 0.12 was used for the B cell, T cell, and yeast datasets, and 0.3 was used for the much larger Drosophila dataset. To estimate the significance of the observations, a permutation test was performed. One-thousand permutations were calculated for each parameter at each time point in the B cell, T cell, and yeast datasets, and 100 permutations were calculated in the D. melanogaster dataset. To be regarded as significant, the p value had to be ≤ 0.05, both in the correlation and permutation calculations.
For TOC, the expression of genes at each time point are separately clustered. For the actual clustering, we used the SOM method, which organized the clusters according to the shapes of the expression profiles of genes. During the method development, we also monitored the clustering process visually; SOMs served as ideal visualization tools. Following TOC, the Spearman correlation was used to test the linear correlations between the medians of gene expression and protein variables in the corresponding clusters.
TOC was an essential step in facilitating the analysis of functional and structural attributes (ontology, subcellular localization, SCOP) because, due to the nature (binary and categorical) and scarceness of some of these data, a clustering measure was required before conducting the correlation test. Furthermore, all correlation coefficients could be compared by unifying the sample sizes with a fixed number of clusters (20) for each of the different types and data samples.
Since the functional and structural information variables were categorical (e.g., a hierarchy of ontology), we conducted correlation tests between the medians of the clusters of expression levels and medians of the ratios (observed/expected in each cluster) of the sequences attributed to the functional or structural variables. The expected number of a given ontology, subcellular localization, or structural variable for each cluster was calculated based on the size of the cluster and the total number of occurrences of the variable.
The correlation coefficients and p values for correlation and permutation tests are in Additional files 1, 2, 3, 4, 5 (CorrelationDataNumbersTCELLCD3.xls, CorrelationDataNumbersTCELLBCDs.xls, CorrelationDataNumbersBCELL.xls, CorrelationDataNumbersYEAST.xls, CorrelationDataNumbersDROSOPHILA.xls) for CD3-induced T cells, CD3/CD28-costimulated T cells, B cells, yeast, and Drosophila datasets, respectively.
Visualization
To provide an intelligible report on the very large correlation analyses, we introduced a new type of visualization. Correlation matrices were formed in which columns and rows represent time points and variables, respectively. Each cell represents a correlation coefficient (theoretically ranging from -1 to 1) between the expression levels of the time point and the values of the variable. Red and green represent positive and negative correlations, respectively. The intensities of the colors are relative to the absolute value of the correlation coefficients. A significant red-colored proteome variable implies an increase in production (i.e., overexpression of proteins either attributed to a property concerning functional or structural variables [GO, SCOP, and SubLoc] or containing higher proportions of a property [amino acids and physicochemical parameters]) compared with the changes in expression of other proteins, which implies the enrichment of the property at that time point. Likewise, a significant green-colored variable implies a reduction in the production of proteins either associated with or containing lower proportions of the property (i.e., underrepresentation of the property at the time point). To determine which variables covary in a similar manner with expression levels over time, the matrix was clustered using SOMs.
Tools and software
Script from CPAN was used for clustering. We developed the software for all other analyses and calculations as well as for all visualizations.
Authors' contributions
MTAS participated in the statistical analysis, preliminary program development, and drafted the manuscript. MS carried out the data mining and statistical analysis, developed computer tools, and drafted the manuscript. TOL participated in data analysis and developed data mining, analysis and visualization tools. MV conceived of the study, participated in its design and coordination, and drafted the manuscript. All authors read and approved the final manuscript.
Supplementary Material
Additional File 1
CD3-stimulated T cells
Click here for file
Additional File 2
CD3/CD28-costimulated T cells
Click here for file
Additional File 3
B cells
Click here for file
Additional File 4
Synchronized yeast cells
Click here for file
Additional File 5
Drosophila life cycle dataset
Click here for file
Acknowledgements
Teemu Kivioja, Petteri Sintonen and Timo Tiirikka are thanked for spirited discussions and helpful comments. We gratefully acknowledge the financial support from the National Technology Agency of Finland and the Medical Research Fund of Tampere University Hospital.
==== Refs
Slonim DK From patterns to pathways: gene expression data analysis comes of age Nat Genet 2002 32 Suppl 502 508 12454645 10.1038/ng1033
Niehrs C Pollet N Synexpression groups in eukaryotes Nature 1999 402 483 487 10591207 10.1038/990025
Cohen BA Mitra RD Hughes JD Church GM A computational analysis of whole-genome expression data reveals chromosomal domains of gene expression Nat Genet 2000 26 183 186 11017073 10.1038/79896
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
Brazma A Jonassen I Vilo J Ukkonen E Predicting gene regulatory elements in silico on a genomic scale Genome Res 1998 8 1202 1215 9847082
Roth FP Hughes JD Estep PW Church GM Finding DNA regulatory motifs within unaligned noncoding sequences clustered by whole-genome mRNA quantitation Nat Biotechnol 1998 16 939 945 9788350
Jansen R Gerstein M Analysis of the yeast transcriptome with structural and functional categories: characterizing highly expressed proteins Nucleic Acids Res 2000 28 1481 1488 10684945 10.1093/nar/28.6.1481
Ollila J Vihinen M Microarray analysis of B-cell stimulation Vitam Horm 2002 64 77 99 11898398
Doniger SW Salomonis N Dahlquist KD Vranizan K Lawlor SC Conklin BR MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data Genome Biol 2003 4 R7 12540299 10.1186/gb-2003-4-1-r7
Draghici S Khatri P Bhavsar P Shah A Krawetz SA Tainsky MA Onto-Tools, the toolkit of the modern biologist: Onto-Express, Onto-Compare, Onto-Design and Onto-Translate Nucleic Acids Res 2003 31 3775 3781 12824416 10.1093/nar/gkg624
Zeeberg BR Feng W Wang G Wang MD Fojo AT Sunshine M Narasimhan S Kane DW Reinhold WC Lababidi S Bussey KJ Riss J Barrett JC Weinstein JN GoMiner: a resource for biological interpretation of genomic and proteomic data Genome Biol 2003 4 R28 12702209 10.1186/gb-2003-4-4-r28
Ge H Liu Z Church GM Vidal M Correlation between transcriptome and interactome mapping data from Saccharomyces cerevisiae Nat Genet 2001 29 482 486 11694880 10.1038/ng776
Kemmeren P van Berkum NL Vilo J Bijma T Donders R Brazma A Holstege FC Protein interaction verification and functional annotation by integrated analysis of genome-scale data Mol Cell 2002 9 1133 1143 12049748 10.1016/S1097-2765(02)00531-2
Fukuchi S Nishikawa K Protein surface amino acid compositions distinctively differ between thermophilic and mesophilic bacteria J Mol Biol 2001 309 835 843 11399062 10.1006/jmbi.2001.4718
Cedano J Aloy P Perez-Pons JA Querol E Relation between amino acid composition and cellular location of proteins J Mol Biol 1997 266 594 600 9067612 10.1006/jmbi.1996.0804
Drawid A Jansen R Gerstein M Genome-wide analysis relating expression level with protein subcellular localization Trends Genet 2000 16 426 430 11050323 10.1016/S0168-9525(00)02108-9
Hegyi H Gerstein M The relationship between protein structure and function: a comprehensive survey with application to the yeast genome J Mol Biol 1999 288 147 164 10329133 10.1006/jmbi.1999.2661
Andrade MA O'Donoghue SI Rost B Adaptation of protein surfaces to subcellular location J Mol Biol 1998 276 517 525 9512720 10.1006/jmbi.1997.1498
Tekaia F Yeramian E Dujon B Amino acid composition of genomes, lifestyles of organisms, and evolutionary trends: a global picture with correspondence analysis Gene 2002 297 51 60 12384285 10.1016/S0378-1119(02)00871-5
Diehn M Alizadeh AA Rando OJ Liu CL Stankunas K Botstein D Crabtree GR Brown PO Genomic expression programs and the integration of the CD28 costimulatory signal in T cell activation Proc Natl Acad Sci U S A 2002 99 11796 11801 12195013 10.1073/pnas.092284399
Ollila J Vihinen M Stimulation-induced gene expression in Ramos B-cells Genes Immun 2003 4 343 350 12847549 10.1038/sj.gene.6363974
Spellman PT Sherlock G Zhang MQ Iyer VR Anders K Eisen MB Brown PO Botstein D Futcher B Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization Mol Biol Cell 1998 9 3273 3297 9843569
Arbeitman MN Furlong EE Imam F Johnson E Null BH Baker BS Krasnow MA Scott MP Davis RW White KP Gene expression during the life cycle of Drosophila melanogaster Science 2002 297 2270 2275 12351791 10.1126/science.1072152
Mustelin T Tasken K Positive and negative regulation of T-cell activation through kinases and phosphatases Biochem J 2003 371 15 27 12485116 10.1042/BJ20021637
Viola A Schroeder S Sakakibara Y Lanzavecchia A T lymphocyte costimulation mediated by reorganization of membrane microdomains Science 1999 283 680 682 9924026 10.1126/science.283.5402.680
Ashburner M Ball CA Blake JA Botstein D Butler H Cherry JM Davis AP Dolinski K Dwight SS Eppig JT Harris MA Hill DP Issel-Tarver L Kasarskis A Lewis S Matese JC Richardson JE Ringwald M Rubin GM Sherlock G Gene ontology: tool for the unification of biology. The Gene Ontology Consortium Nat Genet 2000 25 25 29 10802651 10.1038/75556
Andreeva A Howorth D Brenner SE Hubbard TJ Chothia C Murzin AG SCOP database in 2004: refinements integrate structure and sequence family data Nucleic Acids Res 2004 32 Database issue D226 9 14681400 10.1093/nar/gkh039
Akashi H Gojobori T Metabolic efficiency and amino acid composition in the proteomes of Escherichia coli and Bacillus subtilis Proc Natl Acad Sci U S A 2002 99 3695 3700 11904428 10.1073/pnas.062526999
Greenbaum D Jansen R Gerstein M Analysis of mRNA expression and protein abundance data: an approach for the comparison of the enrichment of features in the cellular population of proteins and transcripts Bioinformatics 2002 18 585 596 12016056 10.1093/bioinformatics/18.4.585
Anderson L Seilhamer J A comparison of selected mRNA and protein abundances in human liver Electrophoresis 1997 18 533 537 9150937 10.1002/elps.1150180333
Gygi SP Rochon Y Franza BR Aebersold R Correlation between protein and mRNA abundance in yeast Mol Cell Biol 1999 19 1720 1730 10022859
Nel AE T-cell activation through the antigen receptor. Part 1: signaling components, signaling pathways, and signal integration at the T-cell antigen receptor synapse J Allergy Clin Immunol 2002 109 758 770 11994696 10.1067/mai.2002.124259
Nel AE Slaughter N T-cell activation through the antigen receptor. Part 2: role of signaling cascades in T-cell differentiation, anergy, immune senescence, and development of immunotherapy J Allergy Clin Immunol 2002 109 901 915 12063516 10.1067/mai.2002.124965
Cho RJ Campbell MJ Winzeler EA Steinmetz L Conway A Wodicka L Wolfsberg TG Gabrielian AE Landsman D Lockhart DJ Davis RW A genome-wide transcriptional analysis of the mitotic cell cycle Mol Cell 1998 2 65 73 9702192 10.1016/S1097-2765(00)80114-8
Gerstein M Jansen R The current excitement in bioinformatics-analysis of whole-genome expression data: how does it relate to protein structure and function? Curr Opin Struct Biol 2000 10 574 584 11042457 10.1016/S0959-440X(00)00134-2
Ruse CI Tan FL Kinter M Bond M Intregrated analysis of the human cardiac transcriptome, proteome and phosphoproteome Proteomics 2004 4 1505 1516 15188417 10.1002/pmic.200300682
McRedmond JP Park SD Reilly DF Coppinger JA Maguire PB Shields DC Fitzgerald DJ Integration of proteomics and genomics in platelets: a profile of platelet proteins and platelet-specific genes Mol Cell Proteomics 2004 3 133 144 14645502 10.1074/mcp.M300063-MCP200
Mijalski T Harder A Halder T Kersten M Horsch M Strom TM Liebscher HV Lottspeich F de Angelis MH Beckers J Identification of coexpressed gene clusters in a comparative analysis of transcriptome and proteome in mouse tissues Proc Natl Acad Sci U S A 2005 102 8621 8626 15939889 10.1073/pnas.0407672102
Ribeiro JM Charlab R Pham VM Garfield M Valenzuela JG An insight into the salivary transcriptome and proteome of the adult female mosquito Culex pipiens quinquefasciatus Insect Biochem Mol Biol 2004 34 543 563 15147756 10.1016/j.ibmb.2004.02.008
Knight CG Kassen R Hebestreit H Rainey PB Global analysis of predicted proteomes: functional adaptation of physical properties Proc Natl Acad Sci U S A 2004 101 8390 8395 15150418 10.1073/pnas.0307270101
Benga G Ferdinand W Amino acid composition of rat and human liver microsomes in normal and pathological conditions Biosci Rep 1995 15 111 116 7579035 10.1007/BF01200145
Forli L Pedersen JI Bjortuft Vatn M Kofstad J Boe J Serum amino acids in relation to nutritional status, lung function and energy intake in patients with advanced pulmonary disease Respir Med 2000 94 868 874 11001078 10.1053/rmed.2000.0830
The FlyBase database of the Drosophila genome projects and community literature Nucleic Acids Res 2003 31 172 175 12519974 10.1093/nar/gkg094
Ikai A Thermostability and aliphatic index of globular proteins J Biochem (Tokyo) 1980 88 1895 1898 7462208
Fisher HF A Limiting Law Relating the Size and Shape of Protein Molecules to Their Composition Proc Natl Acad Sci U S A 1964 51 1285 1291 14215653
Bigelow CC On the average hydrophobicity of proteins and the relation between it and protein structure J Theor Biol 1967 16 187 211 6048539 10.1016/0022-5193(67)90004-5
Waugh DF Protein-protein interactions Adv Protein Chem 1954 9 325 437 13217921
Kyte J Doolittle RF A simple method for displaying the hydropathic character of a protein J Mol Biol 1982 157 105 132 7108955 10.1016/0022-2836(82)90515-0
Eisenberg D Weiss RM Terwilliger TC The helical hydrophobic moment: a measure of the amphiphilicity of a helix Nature 1982 299 371 374 7110359 10.1038/299371a0
Janin J Surface and inside volumes in globular proteins Nature 1979 277 491 492 763335 10.1038/277491a0
Hopp TP Woods KR Prediction of protein antigenic determinants from amino acid sequences Proc Natl Acad Sci U S A 1981 78 3824 3828 6167991
Baumann G Frommel C Sander C Polarity as a criterion in protein design Protein Eng 1989 2 329 334 2928295
Bachmair A Finley D Varshavsky A In vivo half-life of a protein is a function of its amino-terminal residue Science 1986 234 179 186 3018930
Vihinen M Relationship of protein flexibility to thermostability Protein Eng 1987 1 477 480 3508295
Vihinen M Torkkila E Riikonen P Accuracy of protein flexibility predictions Proteins 1994 19 141 149 8090708 10.1002/prot.340190207
Hua S Sun Z Support vector machine approach for protein subcellular localization prediction Bioinformatics 2001 17 721 728 11524373 10.1093/bioinformatics/17.8.721
Krogh A Larsson B von Heijne G Sonnhammer EL Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes J Mol Biol 2001 305 567 580 11152613 10.1006/jmbi.2000.4315
Maglott D Ostell J Pruitt KD Tatusova T Entrez Gene: gene-centered information at NCBI Nucleic Acids Res 2005 33 D54 8 15608257 10.1093/nar/gki031
Gough J Karplus K Hughey R Chothia C Assignment of homology to genome sequences using a library of hidden Markov models that represent all proteins of known structure J Mol Biol 2001 313 903 919 11697912 10.1006/jmbi.2001.5080
McClure MA Smith C Elton P Parameterization studies for the SAM and HMMER methods of hidden Markov model generation Proc Int Conf Intell Syst Mol Biol 1996 4 155 164 8877515
Mulder NJ Apweiler R Attwood TK Bairoch A Barrell D Bateman A Binns D Biswas M Bradley P Bork P Bucher P Copley RR Courcelle E Das U Durbin R Falquet L Fleischmann W Griffiths-Jones S Haft D Harte N Hulo N Kahn D Kanapin A Krestyaninova M Lopez R Letunic I Lonsdale D Silventoinen V Orchard SE Pagni M Peyruc D Ponting CP Selengut JD Servant F Sigrist CJ Vaughan R Zdobnov EM The InterPro Database, 2003 brings increased coverage and new features Nucleic Acids Res 2003 31 315 318 12520011 10.1093/nar/gkg046
|
16131395
|
PMC1236912
|
CC BY
|
2021-01-04 16:27:27
|
no
|
BMC Bioinformatics. 2005 Aug 30; 6:215
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-215
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2191614454810.1186/1471-2105-6-219SoftwareGeneralizations of Markov model to characterize biological sequences Wang Junwen [email protected] Sridhar [email protected] Penn Center for Bioinformatics, Department of Genetics, University of Pennsylvania Philadelphia, PA 19104-6021, USA2005 6 9 2005 6 219 219 29 4 2005 6 9 2005 Copyright © 2005 Wang and Hannenhalli; licensee BioMed Central Ltd.2005Wang and Hannenhalli; 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 currently used kth order Markov models estimate the probability of generating a single nucleotide conditional upon the immediately preceding (gap = 0) k units. However, this neither takes into account the joint dependency of multiple neighboring nucleotides, nor does it consider the long range dependency with gap>0.
Result
We describe a configurable tool to explore generalizations of the standard Markov model. We evaluated whether the sequence classification accuracy can be improved by using an alternative set of model parameters. The evaluation was done on four classes of biological sequences – CpG-poor promoters, all promoters, exons and nucleosome positioning sequences. Using di- and tri-nucleotide as the model unit significantly improved the sequence classification accuracy relative to the standard single nucleotide model. In the case of nucleosome positioning sequences, optimal accuracy was achieved at a gap length of 4. Furthermore in the plot of classification accuracy versus the gap, a periodicity of 10–11 bps was observed which might indicate structural preferences in the nucleosome positioning sequence. The tool is implemented in Java and is available for download at .
Conclusion
Markov modeling is an important component of many sequence analysis tools. We have extended the standard Markov model to incorporate joint and long range dependencies between the sequence elements. The proposed generalizations of the Markov model are likely to improve the overall accuracy of sequence analysis tools.
==== Body
Background
Biological complexity has evolved through a combination and interactions between simpler units. By looking at these units in a context dependent way, we can better understand the biological complexity. For example, Wang and Feng explored the amino acid propensity pattern in a neighbor-dependent way and found that the patterns were not always predictable from the single amino acid patterns [1]. Application of these di-amino acid propensity patterns into a traditional Needleman-Wunsch [2] algorithm significantly improved protein sequence alignment [3]. Similarly one can better predict the transcription factor binding sites by considering the interdependence between nucleotides [4,5].
Markov model (MM) is a statistical technique to model sequences such that the probability of a sequence element is based on a limited context preceding the element [6,7]. In other words, MM is a way to factorize the probability of observing the sequence in terms of context-dependent probabilities of the sequence elements. It has been effectively used in many DNA sequence recognition problems such as promoter and gene prediction [8]. The standard kth order MM assumes that a sequence element probability depends on k previous bases, immediately preceding the current base. Alternatively, the standard Markov Model generates a single base (model unit size = 1) according to a probability distribution depending on the k bases immediately preceding the generated base (gap = 0).
The biological rationale behind selecting these parameters is not clear and alternatives should be explored. Longer range dependencies and joint dependency of neighboring bases have been observed in protein and DNA sequences. For instance, CG di-nucleotide is what characterizes CpG islands [1,9]. In bacterial promoters, a regular positioning of TA and TG stacks is prevalent with the best fit period 5.6 bp [10]. Stacking between neighboring bases is an important source of enthalpy change on helix formation [11]. In the study by Ozoline et al. the period of 5.6 bps for TA and TG can be interpreted as half of the helical repeat period with a contribution from a sequence-dependent helical writhe of the promoter DNA [10]. A repetition of certain di-nucleotide at 10–11 bp has been discovered in numerous genomes, supporting the DNA wrapping around the nucleosomes [12]. A model with unit size of 2 might be more appropriate to characterize the joint dependency of CG di-nucleotide. Furthermore, longer range dependencies (gap > 0) should be explored to model the periodicity of helix pattern. These alternative hypotheses regarding the positional and joint dependence within sequences can be computationally evaluated by extending the Markov Models.
There have been attempts to generalize Markov models. The Mixture Transition Distribution Model conditions the current state on a combination of previous states at varying distances [13]. In the spatial model, the current nucleotide depends on both the left and the right nucleotides [14]. For a detailed review of other generalizations and their limitations see [15]. We have developed a configurable tool to allow for generalizations of Markov model (GMM), as described in the implementation section.
We have evaluated a few instances of our GMM for their ability to classify four classes of sequences – CpG-poor promoters, all promoters, exons and nucleosome positioning sequences against appropriate background sequences. We compared two special cases of our model, the third order di-nucleotide (model unit size = 2) and 2nd order tri-nucleotide (model unit size = 3) GMM against the traditional 6th order single nucleotide Markov model. Our results based on 10-fold cross validation show that the di-nucleotide and the tri-nucleotide based models are significantly better than the single nucleotide based models. Furthermore, in the case of nucleosome positioning sequences, a gap length of 4 achieves the optimal classification accuracy. By allowing us to explore the dependence structure, the GMM tool will not only improve the classification accuracy of a sequence class, but will also provide insights into the structural properties of the sequences.
Implementation
We define the bases whose probability we wish to compute as the posterior bases or simply the posterior and the bases on which this probability is conditioned upon, as the prior. We use six parameters to specify a Markov model (as shown in Figure 1). To capture the joint dependency of neighboring nucleotides, our model allows multiple nucleotides as the model unit. However we allow different model unit sizes for the prior and the posterior, denoted as L1 and L2 respectively. The gap between the posterior and prior is denoted by G. The prior is composed of a few individual model units. The number of such units is called order. The maximum order is denoted by O. We also allow these individual prior units to be at an arbitrary spacing from each other. This spacing between the prior units is denoted by g1. Lastly, within the bases comprising the posterior we allow arbitrary spacing between the bases denoted as g2. For instance a spacing of length 2 within a posterior model unit of size 2 in an amino acid sequence captures the joint dependency for the first and the fourth amino acid residue, which is likely to form a hydrogen bond – vital for the protein helix structure [16]. To evaluate a model where each tri-nucleotide depends on the previous 4 bases, one can set L1 = 4, 0 = 1, L2 = 3, g1 = g2 = G = 0. To use the four bases after ignoring the immediately preceding 3 bases, one can set G = 3.
Figure 1 The figure illustrates the six configurable parameters. There are three parameter associated with the prior – model unit size L1, order (number of units) 0, and spacing between units g1. There are two parameter associated with the posterior – model unit size L2, and spacing between bases g2. And there is a gap parameter G. Although not as general as arbitrary graphical models, this implementation is highly configurable with respect to model unit sizes and the dependence structures in terms of gap lengths.
The prior order O only specifies the maximum order. Our program uses the idea of variable length Markov model [17] such that the highest order for which sufficient data is available, is utilized [18,19]. Apart from the 6 parameters mentioned above, the other generic configurable parameters include: type of biological sequence, either protein ('P') or DNA ('N'); threshold for minimal count of prior for k-mer elimination; pseudo count for a k-mer absent in the training set and the phase the user wants to score. For further information on the parameters, please refer to the software package readme file. Given a particular configuration, our implementation of the GMM is very similar to that of GLIMMER package, with a few exceptions.
Training
In order to achieve statistical robustness, we only consider the k-mers above a (configurable) frequency threshold in the positive sequences. This frequency must ensure that the estimated conditional probabilities are acceptably close to true probabilities. A frequency threshold of 400 was estimated in [19] that provided a 95% confidence that the estimated probabilities were within 0.05 from the true ones. We tried varying this threshold from 50 to 500 and it did not make a significant difference in performance (also observed in [19]) and attains the maximum at 300. Hence we chose this as the default threshold. For nucleosome sequences we chose 50 as the frequency threshold due to smaller data set.
We slide the window one base at a time along the training sequence. The window size is determined by the user defined parameters; window size = L1 × O + g1 × (O-1) + G + L2 + g2 × (L2-1). For each window, we extract the words corresponding to the prior and the posterior. For example, for L1 = 1, O = 6, L2 = 2, g1 = 0, G = 1, g2 = 1, we have a window with length 10, say ACTGATGCAG. The di-nucleotide CG represents the posterior. We increment the counts of k-mers ACTGATCG (6th order), CTGATCG (5th order), ..., and CG (0th order) by one. We thus have 7 sub-models, one for each order.
Once the training sequences are processed, we convert the raw counts into transition probabilities. For the 0th order, the probability is the composition of the L2-mers. For higher order, say, 4th order TGATCG, we compute the sum of frequencies of all the hexamers of the form TGAT**. If the sum is bigger than the user specified threshold, we calculate the probability by dividing the count of TGATCG by the sum. Otherwise, the program automatically uses the (k-1)-mer, and so on to order 0, where the base composition is used. The same process is repeated for the background training sequences and we thus obtain a negative model. We then convert the probability for each k-mer into log-odds score.
Testing
The program first reads in the model – the k-mer log-odds – along with the configuration file. Scoring proceeds in a sliding window fashion where each window is the minimal sequence containing a posterior and the prior as described above. To score a window, we first consider the highest order. Using the example above, to score ACTGATGCAG (the underscored bases correspond to gaps in the model), we first look for 6th order dependence, ie., ACTGATCG in the 8-mer table. If the string exists, we use the score. Otherwise, we look for the string corresponding to the 5th order (CTGATCG), and so on, until the 0th order, ie., the di-nucleotide composition. The sequence score is obtained by adding all window scores. We score the sequence using two models corresponding to the positive and the negative sequences.
For posterior length L2, the overall sequence score can be interpreted as the sum of the scores of L2 independent parses of the sequence in different phases. In each parse or phase, any given base is generated exactly once. We will illustrate this with an example. Let L2 = 3, and g2 = 1. Consider a test sequence S = s1s2...sn. The posterior Pi starting at ith position is sisi+2si+4. Each Pi is in a specific phase φk, 1 ≤ κ ≤ L2. Under φ1 we consider P1,P2,P7,P8,P13,P14, .... We jump from P2 to P7, since all bases from s1 to s6 are covered by P1 and P2. Similarly under φ2 we consider P3,P4,P9,P10,P15,P16, .... Hence the phases for Pi, i = 1,2,3..... are 1,1,2,2,3,3,1,1,2,2,3,3.... Note that each base position is covered exactly once in any of the three phases. If one has a prior knowledge of sequence phase (eg. in-phase exons) and does not wish to use the sum of all phases as a sequence score, one can specify a particular phase to be used. The model will use only the posteriors in that specific phase for training and scoring.
Results
The human promoter sequences
We extracted the ± 5 kb region around the 12,333 Transcriptional Start Site (TSS) in the DBTSS database [20]. These start sites are identified using oligo-capping approach. We have implemented a sliding-window based program to identify CpG-islands using the original definition of CpG-islands [21]. We have also implemented a Hidden Markov Model (HMM) approach for CpG island identification [7]. We call a 10 kb promoter region CpG-poor if it does not contain a 200 bp length CpG-island by either of the two methods. This resulted in 1,466 CpG-poor promoters from a total of 12,333 promoter sequences. We then randomly selected 5,000 10 kb sequences from the whole human genome after masking the DBTSS promoter regions. The 5,000 background sequences and the 1,466 CpG-poor promoters were used to evaluate the various models. The same background dataset was also used for the classification of the entire set of 12,333 promoter sequences.
The human exon dataset
The human exon locations were downloaded from UCSC genome browser, human genome version hg16. We extracted the exon sequences based on start and end locations. We thus obtained 219,624 exons. To compile a background sequence set, we randomly selected the same length sequences from the background for each exon.
The nucleosome positioning sequences
The nucleosome positioning sequences were downloaded from the Nucleosome Positioning Region Database (NPRD) [22, 31]. The generation of background sequences was done similarly to the exon dataset.
Model evaluation
We used 10-fold cross-validations to train and test the models. The positive and the background sequences were randomly partitioned into 10 equal parts. Each part was tested after training on the other 9 parts. Once the models were trained, we scored the training set using the models and obtained a cutoff based on the specificity-sensitivity curve. We chose a score cutoff that resulted in the best correlation coefficient (CC) value for the training set. We then scored the (independent) test set and applied this cutoff to obtain the CC value. The mean and standard deviation over the 10 CC values was calculated. The Sensitivity (Se), Specificity (Sp) and Correlation coefficient (CC) values were defined as following:
TP: True positive, FP: False positive, TN: True negative, FN: False negative.
We have provided scripts to evaluate a specified configuration based on 10-fold cross validation. This involves scripts for splitting the input sequence into 10 equal parts and code for calculating the sensitivity, specificity and correlation coefficient.
To assess the significance of the performance improvement using a model M compared to base model M* (standard MM), we used Wilcoxon paired rank sum test. All sequences (positive and background) were scored using M to obtain score list S and using M* to obtain score list S*. Both S and S* were normalized separately to mean 0 and standard deviation 1. These paired normalized scores for positive sequences (each sequence has 2 scores corresponding to the 2 models) were used to test whether the scores in S* are greater than the corresponding scores in S using Wilcoxon test.
We have applied specific configurations of the tool to a few biological sequence classification problems as an illustration. Specifically to evaluate the impact of varying model unit size we used the following three settings:
(1) 6th order single nucleotide model: L1 = L2 = 1, O = 6, g1 = 0, G = 0, g2 = 0,
(2) 3rd order di-nucleotide model: L1 = L2 = 2, O = 3, g1 = 0, G = 0, g2 = 0,
(3) 2th order tri-nucleotide model: L1 = L2 = 3, O = 2, g1 = 0, G = 0, g2 = 0.
The 6th order single-nucleotide Markov Model is common in many sequence analysis tools currently used. Notice that the total number of prior bases is six for each of these three models. We tested the classification accuracy for three sequence classes using the above three configurations. The results for CpG-poor promotes, all promoters and all exon classifications are showed in Table 1, and discussed below.
Table 1 Average and standard deviation of Correlation coefficient (CC) values using different models. The data were obtained from 10 cross-validation. The CC values were obtained from testing dataset when cutoff selected from the training set. * Wilcox rank sum paired test shows significant (p-value < 0.001) better than the corresponding single nucleotide model.
Samples (size) Single nucleotide Di-nucleotide Tri-nucleotide
CpG-poor Promoters (1,466) 0.24 ± 0.05 0.28 ± 0.03* 0.34 ± 0.04*
All Promoters (12,333) 0.54 ± 0.02 0.54 ± 0.03 0.56 ± 0.02*
All Exons (219,624) 0.63 ± 0.00 0.64 ± 0.00* 0.67 ± 0.00*
Classification of CpG-poor promoters
The di- and tri- nucleotide models improve upon single nucleotide model (p-value < 0.001). The traditional (single-nucleotide) 6th order Markov model yielded a correlation coefficient value (CC) of 0.24. When we use the tri- nucleotide model, the CC value was improved by 39% to 0.34. The specificity-sensitivity graph (Figure 2) further shows the sensitivity-specificity tradeoff. For instance, at a reference sensitivity value of 0.5, the specificity achieved by the tri-nucleotide based model is 0.52, as compared to 0.36 for the single-nucleotide model.
Figure 2 The specificity-sensitivity graph of the discrimination of CpG-poor promoters against background sequences using three different models – 6th order single nucleotide model (Red), 3rd order di-nucleotide model (Blue), and 2nd order tri-nucleotide model (Black).
Classification of all promoters
We next applied the models to classify the entire set of 12,333 promoter sequences. The tri-nucleotide model shows an improvement in the classification accuracy (0.58 versus 0.54, p-value < 0.001) relative to the single nucleotide model. The entire set of promoters is dominated by CpG associated promoters which by virtue of being GC-rich and containing CpG islands have a strongly distinguishable characteristics against the background sequences. Consequently the relative gains of using larger model units are marginal.
Classification of exons
We extracted 219,624 annotated exons from the hg16. We randomly selected the 219,624 sequences with the same length as the exons from background sequences. The average correlation coefficient for classification accuracy for single-, di-, and tri-nucleotide models are 0.63, 0.645 and 0.66 respectively. This modest improvement is however statistically significant.
Classification of nucleosome positioning sequences
A periodical distribution pattern of transcription factor sites was observed in promoter region that suggested a correlation between the positioning of nucleosomes and transcription factor binding sites [23]. To investigate the nucleosome sequence periodicity, we compared classification accuracy at different gap length (parameter G) between prior and posterior. We were able to obtain 112 nucleosome sequences and performed their classification based on the first order tri-nucleotide model (L1 = 3, O = 1, g1 = 0, L2 = 3, g2 = 0) at varying values of G. We achieve the best classification accuracies at G = 4, 15 and 25, and worst classification accuracies at G = 7 and 18 (Figure 3). The distances between consecutive peaks and valleys are around 10–11 bps, which is close to DNA helix turn of 10.5 bps (for most common B-DNA, 11 bps for A-DNA, 12 bps for Z-DNA).
Figure 3 Plot of classification accuracy for the Nucleosome positioning sequences with respect to the gap between the prior and the posterior. This is based on the first order tri-nucleotide model (L1 = 3, O = 1, g1 = 0, L2 = 3, g2 = 0) at varying values of G. We achieve the best classification accuracies (peaks) at G = 4, 15 and 25, and worst classification accuracies (valleys) at G = 7 and 18. The distances between consecutive peaks and valleys are around 10–11 bps, which is close to DNA helix turn of 10.5 bps (for most common B-DNA, 11 bps for A-DNA, 12 bps for Z-DNA). This result illustrates the utility of the tool in exploring such long-range dependencies which might indicate specific structural constraints of the sequence class.
Run time
We compared the run time for the three models on training and testing of the CpG-poor promoter classification against the background. The benchmark was based on 64 Mb sequences with parameters described in the method section. The java program was tested on a 2.6 GHz Pentium III dual processors with 16GB of RAM running linux. The training time for single-nucleotide based model was 55.8 minutes. This reduced to 23.8 and 18.9 minutes for the di- and tri-nucleotide based models respectively. The time needed for testing reduces less significantly by 30%-40%, from 22.9 minutes for single to 15.4 and 14.0 minutes for di- and tri nucleotide models respectively. The run time reductions are mainly due to fewer orders the model needs to go though for di- and tri- nucleotide models.
Discussion
Markov chains are commonly used to model biological sequences. However the specific model unit size and the dependence structure among the sequence elements have not been explored. Specifically the model unit is fixed as a single nucleotide or amino acid and its dependence on k elements immediately preceding the current element is incorporated in the model. We have argued that it might be better to consider different model unit size and dependence structures. Furthermore, it has been reported that the optimal choice of model type and the model order is species specific [18]. Hence, it is important to implement the modeling tool in a configurable fashion.
Promoter prediction
Despite numerous efforts in promoter prediction, the subclass of promoters not associated with CpG islands or CpG-poor promoters are notoriously difficult to characterize and predict. This remains the main bottleneck in overall promoter prediction accuracy and an accurate analysis of transcriptional regulation [24-26]. One component of promoter prediction is a better characterization of overall DNA structural feature in the vicinity of the promoters. Consistent with other studies that by considering the neighboring dependency of amino acids can improve the protein sequence alignment [3], here we show that by using the longer Markov unit to capture the joint dependency of neighboring nucleotides, we can substantially improve the CpG-poor promoter classification. Although we are not proposing an improved promoter prediction tool here, our result does suggests an alternative modeling of the long range DNA characteristics which is likely to improve the overall promoter prediction.
Nucleosome positioning (NP) sequence prediction
The nucleosome is the basic unit of chromatin. Regulation of eukaryotic gene transcription is closely linked with the changes in nucleosome structure of the chromatin [22]. A nucleosome at the promoter region is capable of inhibiting the transcription initiation, whereas its displacement is capable of surmounting the repressive effect [27]. The preference of various sequences to allow for NP is not clear. An interesting aspect of our application of GMM to NP sequences is the observation that a gap length of 4 better captures the local dependence in these sequences. This, along with the periodicity of 10–11 bps in the plot of classification accuracy against the gap length might indicate a structural requirement in protein-DNA interaction in the NP.
Generalizations of MM
Two main challenges in generalizing Markov models are (i) ensuring that the score of a sequence given the model can be appropriately factorized in terms of individual model unit scores (each base is included in exactly one model unit, modulo the edge effects), and (ii) accurate parameter estimation. We have shown that the sequence score can be interpreted as sum of scores using L2 independent parses of the sequence, where L2 is the number of posterior bases. Score of the sequence for each phase can indeed be factorized in terms of scores of disjoint posteriors. However with respect to accurate parameter estimation we have adopted a simple strategy analogous to that for standard MM and the parameter estimation methods developed in [15] may provide more accurate models.
We have used the sum of scores in different phases as the overall sequence score. Using the maximum score among all phases presents another alternative, which might be appropriate for coding exons where the codon impose a phase. When we do not have such a priori knowledge, then using maximum among phase scores may be inappropriate. Also it can be computationally prohibitive since one will need to build separate model of each phase and when scoring a sequence, try all models for all phase structure of the sequence. Thus the computational time for scoring a sequence is L2 *L2 – fold greater than the phase-less scoring. In our current implementation, for the cases where there is a prior knowledge of phase, users can specify a phase parameter, such that the model is built for a specific phase and also applied to the same phase.
Conclusion
We have developed a configurable tool to explore generalizations of Markov models incorporating joint and long range dependencies of the sequence elements. As an illustration, we have shown that by using longer k-mer as Markov model units and specific gap lengths, one can improve the classification accuracy for a variety of biologically important sequence classes. Various tools to predict biological sequences like promoters and genes exploit multiple sequence based characteristics. The long range DNA characteristics are commonly captured using Markov models, eg Genscan [28] and HMMgene [29]. An improvement in this aspect of the prediction has direct implications on overall prediction accuracy of these tools. A complete theoretical development of generalizations of Markov models will require further research. The proposed software provides a means to explore dependency structures for a novel sequence class.
Availability and requirements
The software will be freely available for download . The program requires java version 1.4.2 or above to run, and it is platform independent. Please refer to software package for detailed instruction on how to run the programs.
List of abbreviations used
MM – Markov model
GMM – Generalizations of markov model
TSS – Transcription start site
DBTSS – Data base of transcription start sites
NP – Nucleosome positioning
NPRD – Nucleosome positioning region database
CC – Correlation coefficient
Authors' contributions
JW and SH were involved in the developing the idea and writing the manuscript. JW implemented the software.
Acknowledgements
The authors would like to thank Joan Gu in the lab for compiling part of the dataset. We thank Dr. Yutaka Suzuki for kindly providing us the full-length promoter sequences and Bill Majoros for assistance on the GLIMMER package. Special thanks to reviewers for thoughtful comments.
==== Refs
Wang J Feng JA Exploring the sequence patterns in the alpha-helices of proteins Protein Eng 2003 16 799 807 14631069 10.1093/protein/gzg101
Needleman SB Wunsch CD A general method applicable to the search for similarities in the amino acid sequence of two proteins J Mol Biol 1970 48 443 453 5420325 10.1016/0022-2836(70)90057-4
Wang J Feng JA NdPASA: a novel pairwise protein sequence alignment algorithm that incorporates neighbor-dependent amino acid propensities Proteins 2005 58 628 637 15616964 10.1002/prot.20359
Bulyk ML Johnson PL Church GM Nucleotides of transcription factor binding sites exert interdependent effects on the binding affinities of transcription factors Nucleic Acids Res 2002 30 1255 1261 11861919 10.1093/nar/30.5.1255
Zhou Q Liu JS Modeling within-motif dependence for transcription factor binding site predictions Bioinformatics 2004 20 909 916 14751969 10.1093/bioinformatics/bth006
Davis MHA Markov Models & Optimization Monographs on statistics and applied probability 1993 49 , CHAPMAN & HALL
Durbin R Eddy S Krogh A Mitchison G Biological Sequence Analysis 1998
Ohler U Niemann H Identification and analysis of eukaryotic promoters: recent computational approaches Trends Genet 2001 17 56 60 11173099 10.1016/S0168-9525(00)02174-0
Gardiner-Garden M Frommer M CpG islands in vertebrate genomes J Mol Biol 1987 196 261 282 3656447 10.1016/0022-2836(87)90689-9
Ozoline ON Deev AA Trifonov EN DNA bendability--a novel feature in E. coli promoter recognition J Biomol Struct Dyn 1999 16 825 831 10217452
Dimitrov RA Zuker M Prediction of hybridization and melting for double-stranded nucleic acids Biophys J 2004 87 215 226 15240459 10.1529/biophysj.103.020743
Schieg P Herzel H Periodicities of 10-11bp as indicators of the supercoiled state of genomic DNA J Mol Biol 2004 343 891 901 15476808 10.1016/j.jmb.2004.08.068
Raftery AE A model for high order Markov chains J Roy Statst Soc Ser 1985 B47 528 539
Berchtold A Estimation in the mixture transition distribution model J TIme Ser Anal 2001 22 379 397 10.1111/1467-9892.00231
Berchtold A Raftery AE The mixture transition distribution model for high-order Markov chains and non-Gaussian time series Statistical Science 2002 17 328 356 10.1214/ss/1042727943
Penel S Morrison RG Mortishire-Smith RJ Doig AJ Periodicity in alpha-helix lengths and C-capping preferences J Mol Biol 1999 293 1211 1219 10547296 10.1006/jmbi.1999.3206
Buhlmann P Wyner AJ Variable length markov chains The Annals of Statistics 1999 27 480 513 10.1214/aos/1018031204
Azad RK Borodovsky M Effects of choice of DNA sequence model structure on gene identification accuracy Bioinformatics 2004 20 993 1005 14751980 10.1093/bioinformatics/bth028
Salzberg SL Delcher AL Kasif S White O Microbial gene identification using interpolated Markov models Nucleic Acids Res 1998 26 544 548 9421513 10.1093/nar/26.2.544
Suzuki Y Yamashita R Nakai K Sugano S DBTSS: DataBase of human Transcriptional Start Sites and full-length cDNAs Nucleic Acids Res 2002 30 328 331 11752328 10.1093/nar/30.1.328
Antequera F Bird A Number of CpG islands and genes in human and mouse Proc Natl Acad Sci U S A 1993 90 11995 11999 7505451
Levitsky VG Katokhin AV Podkolodnaya OA Furman DP Kolchanov NA NPRD: Nucleosome Positioning Region Database Nucleic Acids Res 2005 33 (Database Issue) D67 70 15608285
Ioshikhes I Trifonov EN Zhang MQ Periodical distribution of transcription factor sites in promoter regions and connection with chromatin structure Proc Natl Acad Sci U S A 1999 96 2891 2895 10077607 10.1073/pnas.96.6.2891
Davuluri RV Grosse I Zhang MQ Computational identification of promoters and first exons in the human genome Nat Genet 2001 29 412 417 11726928 10.1038/ng780
Hannenhalli S Levy S Promoter prediction in the human genome Bioinformatics 2001 17 Suppl 1 S90 6 11472997
Bajic VB Tan SL Suzuki Y Sugano S Promoter prediction analysis on the whole human genome Nat Biotechnol 2004 22 1467 1473 15529174 10.1038/nbt1032
Li G Chandler SP Wolffe AP Hall TC Architectural specificity in chromatin structure at the TATA box in vivo: nucleosome displacement upon beta-phaseolin gene activation Proc Natl Acad Sci U S A 1998 95 4772 4777 9539814 10.1073/pnas.95.8.4772
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
Krogh A Two methods for improving performance of an HMM and their application for gene finding Proc Int Conf Intell Syst Mol Biol 1997 5 179 186 9322033
|
16144548
|
PMC1236913
|
CC BY
|
2021-01-04 16:27:28
|
no
|
BMC Bioinformatics. 2005 Sep 6; 6:219
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-219
|
oa_comm
|
==== Front
BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1081611149810.1186/1471-2407-5-108Research ArticleThe development of common data elements for a multi-institute prostate cancer tissue bank: The Cooperative Prostate Cancer Tissue Resource (CPCTR) experience Patel Ashokkumar A [email protected] André [email protected] Jules J [email protected] Maarten [email protected] Milton W [email protected] Rajiv [email protected] John [email protected] Jonathan [email protected] Jan [email protected] Kuei-Fang [email protected] Michael J [email protected] Department of Pathology, Center for Pathology Informatics, Benedum Oncology Informatics Center, University of Pittsburgh, Pittsburgh, PA, USA2 Department of Pathology, University of Illinois-Chicago, Chicago, IL, USA3 Cancer Diagnosis Program, National Cancer Institute, Bethesda, MD, USA4 Department of Pathology, New York University, New York, NY, USA5 Departments of Environmental Medicine and Urology, New York University, New York, NY, USA6 Departments of Pathology and Urology, Emory University, Atlanta, GA, USA7 Department of Pathology, George Washington University, Washington, DC, USA8 Bioinformatics Program, Human and Molecular Genetics Center, Medical College of Wisconsin, Milwaukee, WI, USA2005 21 8 2005 5 108 108 24 12 2004 21 8 2005 Copyright © 2005 Patel 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 Cooperative Prostate Cancer Tissue Resource (CPCTR) is a consortium of four geographically dispersed institutions that are funded by the U.S. National Cancer Institute (NCI) to provide clinically annotated prostate cancer tissue samples to researchers. To facilitate this effort, it was critical to arrive at agreed upon common data elements (CDEs) that could be used to collect demographic, pathologic, treatment and clinical outcome data.
Methods
The CPCTR investigators convened a CDE curation subcommittee to develop and implement CDEs for the annotation of collected prostate tissues. The draft CDEs were refined and progressively annotated to make them ISO 11179 compliant. The CDEs were implemented in the CPCTR database and tested using software query tools developed by the investigators.
Results
By collaborative consensus the CPCTR CDE subcommittee developed 145 data elements to annotate the tissue samples collected. These included for each case: 1) demographic data, 2) clinical history, 3) pathology specimen level elements to describe the staging, grading and other characteristics of individual surgical pathology cases, 4) tissue block level annotation critical to managing a virtual inventory of cases and facilitating case selection, and 5) clinical outcome data including treatment, recurrence and vital status. These elements have been used successfully to respond to over 60 requests by end-users for tissue, including paraffin blocks from cases with 5 to 10 years of follow up, tissue microarrays (TMAs), as well as frozen tissue collected prospectively for genomic profiling and genetic studies. The CPCTR CDEs have been fully implemented in two major tissue banks and have been shared with dozens of other tissue banking efforts.
Conclusion
The freely available CDEs developed by the CPCTR are robust, based on "best practices" for tissue resources, and are ISO 11179 compliant. The process for CDE development described in this manuscript provides a framework model for other organ sites and has been used as a model for breast and melanoma tissue banking efforts.
==== Body
Background
Since the completion of the human genome project, there has been a paradigm shift in the way biorepositories have been utilized. Recent advances in the fields of genomics and proteomics are providing novel ways of producing experimental data using biospecimens. This shift has lead to the development of robust clinical annotations for the collected tissues, which easily allows comparative research and in-depth analysis of data among multiple institutions. This new paradigm is further exemplified in 2003 by the RAND Corporation's report on Human Tissue Repositories that recommended "...the collection of consistent and high-quality data associated with every biospecimen and employing a standardized set of common data elements...;" for annotation as a best practice [1].
Common data elements (CDEs) are annotations that are collected in a uniform manner across multiple institutions that allow sharing of data in a standardized format and are defined in detail using a metadata dictionary.
In 1999 the National Cancer Institute (NCI), recognizing the need for a multi-center effort in prostate cancer tissue banking, issued an RFA for a consortium effort to collect large numbers of clinically annotated prostate cancer specimens for the research community [2]. This initiative was created after a similar successful NCI Resource that was created for breast tissue called the Cooperative Breast Cancer Tissue Resource (CBCTR) [3,4]. In April 2000, four academic institutions were funded by the NCI to form a national prostate cancer tissue resource, the Cooperative Prostate Cancer Tissue Resource (CPCTR). The goal of the CPCTR is to collect large numbers of prostate cancer specimens with accurate quality controlled and standardized pathologic review of specimens and detailed, quality controlled outcome data for use in biomarker validation studies, and to make this collection available to the research community. During the initial phase of this project, many of the experiences and basic infrastructural components of the CBCTR were used as a model in developing the CPCTR program. Specifically, the CBCTR data elements were used by the CPCTR team to create common data elements (CDEs) for annotating the archival paraffin embedded tissue samples in the prostate resource; the CBCTR does not include frozen tissue collection.
The process of developing CDEs typically involves many individuals and can take up to several months to arrive at a draft that is based on complete consensus among those involved. In the case of CPCTR there were pathologists, urologists, cancer registrars, data managers, and cancer researchers from five major medical centers and the NCI Cancer Diagnosis Program who provided input and approved changes to the developing CDEs along the process of adopting the initial version. In this process it was essential to 1) include experts from multiple disciplines, 2) consider the works of others creating similar CDEs, and/or 3) consider established standards when available. This communication describes the process of developing CDEs for prostate cancer tissues that are banked by the NCI's CPCTR [5,6].
Methods
Participating institutions
The Resource comprises four academic institutions: George Washington University Medical Center (GWU), Washington, DC; Medical College of Wisconsin (MCW), Milwaukee, WI; New York University School of Medicine (NYU), New York, NY; the University of Pittsburgh (PITT), PA. The Resource has access to cases from a variety of medical care settings that include academic medical centers, as well as private, public, and Veterans Administration hospitals. The participating hospitals are distributed across six states in the Northeastern and Midwestern regions of the US. This varied access to cases allows accrual of cases that reflect a wide diversity of patients undergoing prostate cancer management in the United States.
Human subjects protections
The CPCTR uses a decentralized sample and data collection and storage with a centralized data management repository model. Each CPCTR institution has developed its own local protocols with including consent language describing its procedure to protect the confidentiality and privacy of human subjects and has obtained local IRB approval for all CPCTR activities. Tissue data records from the cooperating institutions are submitted to a central data manager (Information Management Services, Inc. (IMS, Bethesda, MD) [16], contracted by the NCI). All institutions assign a random, ten-digit number generated by IMS to each record before submitting the data to the central database. The only linkage to patient identity is retained locally at each CPCTR institution. This ensures that the central database has no links connecting records to patients. In addition, de-identified datasets are generated from the central database for the research community to query (the so-called safe harbor approach to HIPAA-compliance)
[7]. The ranges of ages are provided instead of the date of birth and diagnosis to meet the compliant requirements and research purposes.
Organization of the Resource
The CPCTR is governed by a Coordinating Committee that has delegated tasks to several sub-committees. Figure 1 describes the three sub-committees that are involved in the CDE development. The Coordinating Committee includes the four principal investigators (PIs), four co-PIs, a biostatistician, the NCI program leader, two central database coordinators, and a member of the Research Evaluation Panel (REP) that reviews request for tissues and data by end-users. The Committee's main function is to oversee all of the activities of the CPCTR. The Committee's role in developing the CDEs was to determine the types of biospecimens (i.e. paraffin archival tissue, frozen tissue, TMAs) the CPCTR will provide. The details of the CPCTR organization are described in the Resource Manual of Operations located at one of the Resource's websites [8].
The pathology sub-committee includes at least one genitourinary (GU) pathologist per member institution, a NCI program leader, and ad hoc urologists and/or prostate cancer researchers. Their major role is to develop standard evaluation guidelines and propose pathology-specific CDEs related to the different types biospecimens collected to the CDE sub-committee.
The data manager sub-committee includes data managers and cancer registrars from each of the four member institutions as well as the CPCTR biostatistician, two central database coordinators, and the NCI program leader. This sub-committee's main role is to implement and evaluate the CDEs and to perform quality assurance checks on the data collected at each member institution and to help coordinate the distribution of tissue requests and the associated data sets.
The CDE sub-committee includes multiple members from each of the previously mentioned committees. This sub-committee's role was to develop CDEs described in the following section.
Development of the Common Data Elements
With guidance from the Coordinating Committee and the other sub-committees, the CDE sub-committee's primary tasked was to develop CDEs for demographics and clinical history, specimen level annotation describing the overall case where a bio-specimen was collected, block level annotation which records information on individual pieces or sections of the bio-specimen banked, and follow-up information about treatment, vital status, biochemical (prostate specific antigen [PSA] values) and clinical recurrence to be included in the database. While utilizing and learning from the experiences of several others groups, the CDE sub-committee particularly took the experiences of the CBCTR into considerations [3,4]. The sub-committee also considered established open source standards including the AJCC Cancer Staging Manual [9], the NAACCR Data Standards for Cancer Registries [10], the CAP Cancer Checklist [11], and other prostate specific CDEs that were available through the NCI Center for Bioinformatics (NCICB) [12,13].
Development of metadata for CDEs
Metadata is additional data developed to describes a specific CDE by following the ISO 11179 standard, which "specifies that metadata should have a qualified name or identifier, an authority who registers the name, a versioning history (allowing for modifications), a language or origin, a statement relating to usage, a data typing statement, and a definition that is unambiguous [14]." The CPCTR data dictionary describing each of the common data elements was generated by following the ISO-11179 standard for meta-data. The most current version of the CDE data dictionary can be accessed at the CPCTR public database website [15]. This version of the documentation was generated by the implementation of Oracle's Application Server (v9.0.2) on a Compaq DL360 Server running Windows 2000 with Service Pack 2. The application, as shown on figure 2, uses the Oracle http server and mod_plsql extensions to generate dynamic pages from the database to the users.
Development of the CPCTR database
Once the initial set of CDEs was developed and approved by the Coordinating Committee, it was used to create a Microsoft Access database by IMS [16]. This database was then distributed to allow each of the member institutes to capture data on all of the tissue samples they will provide to the Resource. Each member institute either utilized this Access database or developed its own database based on it utilizing the technologies that fit its own institutional development environment. Two of the member institutions created their own database using Oracle, while the two other institutions modified the Access database to collect other data elements unique to their local biospecimen collection efforts. By following each institutional IRB's approved protocol, data were collected in the local database. Although, they use different databases, data for all the CDEs from each case were exported to the IMS central database on a monthly basis in pre-defined formats in excel worksheets with the IMS identifier that was randomly generated and pre-assigned to each institution. Once imported into the central database, data QA checks are conducted to detect any missing essential CDEs or possible data input errors including field and cross-field checking (i.e. number of nodes positive >1, then pathology nodal stage = pN1). The valid field options for each data element are defined in the CDE description. Any records with invalid or discrepant data items are censored (i.e., removed from the available tissue samples for investigators) so they are not selected for an application request until they are resolved. Resolutions are the responsibility of the sending institution and are repaired and re-sent with the next monthly data update. In addition, after the initial implementation of any newly created CDEs, there is a short pilot phase where the IMS data check quality assurance review catches any errors and notifies the local institute to resolve the problem.
Furthermore, all of the HIPAA's proscribed set of 18 data elements was omitted from sample records to create a public database [17] for the research community to use. The de-identified data were also utilized by the IMS for filling tissue disbursement to end-users.
Evaluation of the CDEs
The evaluation phase of the CDEs allowed the CPCTR to examine the quality of data collected by each of the member institutions. The evaluation, an ongoing effort, is carried out at multiple levels by the data manager's sub-committee and reported to the Coordinating Committee. As previously mentioned, the initial evaluation is conducted once any specific CDE is approved and changes to Access database are made. A test export file is sent to IMS to verify correct implementation of any new updates. IMS performs a data check for accuracy and completeness and notifies the local institution of any issues to resolve. The second evaluation, conducted on a monthly basis, is the one carried out by the Data manager's sub-committee (QA checks on 10% of all new cases entered into their local database and sent to IMS).
Finally, the data managers also re-evaluated all CDEs in the entire central database once an initial benchmark of 2,000 cases submitted into the central database was reached. The CDE sub-committee's tasks in this final effort were to determine which CDEs were least populated with valid values (e.g., the CDE "patient's history of other cancers"), which valid values were least used to populate a particular CDE (e.g., the CDE "vital status" has valid values of 'alive, alive with prostate cancer, dead, dead with prostate cancer, dead with autopsy, dead with warm autopsy'), and which CDEs created difficulties in collection (e.g., distant site 1 at the time of diagnosis, date of 1st recurrence, 1st non-prostate recurrence, and distant site of 1st recurrence). Standards for the discontinuation, consolidation, or expansion of CDEs were decided on by committee consensus after discussion and review. Individual institutions could choose to keep discontinued CDEs locally if they were associated with specific institutional research goals.
Results
Inventory of resources for CPCTR
At each institution, archival specimens from radical prostatectomies, diagnostic needle biopsies, and surgically removed metastatic tissue specimens from 1989 to present were identified from the pathology records. At the time of compilation of this manuscript (October 2004), there were more than 6,000 annotated cases of prostate cancer specimens with data on over 30,000 archival paraffin-embedded tissues blocks and 5,800 frozen tissue blocks, and two sets of tissue microarrays (TMAs) that are currently available to the research community. The majority of these cases consisted of archival paraffin blocks from surgical patients treated between 1989 and 1998. The remaining cases are recent cases (accrued from 1999 onwards) with prospectively banked tissue (both frozen and paraffin embedded tissue). At some CPCTR sites (e.g. Medical College of Wisconsin, George Washington University, and University of Pittsburgh), blood, serum, and urine samples have also been collected prior to or at the time of surgery from prospectively banked radical prostatectomy patients. The Resource has also accrued diagnostic needle biopsy specimens from at least 2,209 of the radical prostatectomy patients that are entered into the Resource and from 940 prostate cancer patients who did not undergo a radical prostatectomy. The latter samples represent patients who were not eligible for prostatectomy, and received radiation or hormonal therapy, underwent watchful waiting or have died from other causes including other cancers.
Development of the CDEs
The Coordinating Committee [18] created four main data categories for annotation of the types of specimens banked as a guideline for the CDE development process. The four main categories were: 1) Patient demographics and clinical history data; 2) Specimen annotation, which records basic overall information on a particular event where a bio-specimen was collected as a result of a clinical intervention and/or a specific banking event for research based on a protocol; 3) Block level annotation which records attributes detailing each specimen's paraffin or frozen tissue block entered into the Resource (so-called "matrix blocks") from a particular case; and 4) Treatment and outcomes annotation, which records data that is collected in a longitudinal manner through an "event table" so that outcomes based research can be performed. Sub-categories and additional data elements for each of the four main groups are described in figure 3.
The CDE sub-committee identified and developed CDEs within each of the four main categories by reviewing data elements created by other existing consortiums and open sources. Specifically, the CDE sub-committee used the CBCTR core model and expanded it to include detailed block level annotation, including multiple types of tissue [i.e. prostatectomy (frozen and paraffin), biopsy, lymph node, metastasis]. In addition, the CDE sub-committee developed a set of data elements to address clinical outcome by expanding treatment and recurrence fields to include initiation and completion dates so that a clinical timeline of major events can be followed over time for a patient's course of disease. These annotations importantly include the prostate tumor serum marker (PSA) critical to determining biochemical recurrence. The group also used established standards from the American Joint Commission on Cancer (AJCC) for the staging data elements, the College of American Pathologists (CAP) check list for the annotation at a specimen level, and some elements from the North American Association of Central Cancer Registries (NAACCR) for demographics and follow up data elements [9-11].
Common Data Elements
The Resource database consists of a set of 145 common data elements (CDEs) which capture the clinical, pathologic and tissue sample inventory data for each case. The data set for the biopsy and metastatic specimens from patients who did not undergo surgery includes many of the radical prostatectomy CDEs with slight variations on individual block descriptors. Each of the data elements is fully described as a set of features conforming to the ISO-11179 standard for meta-data [14]. The data dictionary detailing these CDEs and the associated paper data forms, used by the pathologists and data managers for capturing data, is included as an attachment to this article. The most current version of the approved CDEs can be found on the CPCTR public database website [15].
In order to facilitate material tracking and the identification of specimen-specific characteristics needed during tissue send out or processing, the Resource captures the pathology characteristics of the tissue specimens at the level of individual paraffin tissue blocks and frozen tissue slices for each case. These annotations were named "block matrix" because the annotations are spreadsheet-like in character. From each case the "block matrix" was applied to paraffin blocks of cancer that were selected for inclusion in the Resource collection, with annotation for each block of histologic tumor type, size of tumor focus, Gleason grade, presence of high grade prostatic intraepithelial neoplasia (HGPIN), and presence of perineural, seminal vesicle, or lymphovascular invasion. A detailed annotation is also used for a lymph node matrix block of cases where metastatic tumor was identified in lymph nodes removed at the time of radical prostatectomy. The two other types of matrix blocks annotated by the Resource are blocks containing HGPIN (but no cancer), and benign prostate tissue from areas adjacent to tumor containing prostatectomy blocks, which may be used as one form of "control" tissue.
The CDE sub-committee also determined that 14 of the CDEs developed were critical (or 'required') data fields, listed on figure 4. These critical data items are the minimum data elements required for a case to be eligible for inclusion in the Resource. Any records with missing or invalid critical data items are rejected and the corresponding institution is responsible for resolving the issue. In additional, there are 17 "conditional" required data items.
Re-evaluation of CDEs
Annual re-evaluation of the CDEs by the data managers of the CPCTR is conducted to determine which CDEs are most useful for routine tissue and data collection and for long-term updates by all the sites. An overview of the process is described in figure 5. If any desired CDE was found to be poorly collected from a quality control or practical standpoint, discussions were initiated to modify the data collection process through discussions with the cancer registrars and data managers. For example, the evaluation of the data collected for recurrence and progression showed that the initial definitions for CDEs caused difficulties for the data managers collecting this information prompting re-definition of those CDEs. The definitions of the CDEs and their metadata were found to be critical in the clear understanding of what information was to be collected. For example, the initial collection of data related to the CPCTR's CDE for distant metastasis and recurrences were entered in multiple fields (i.e., distant metastasis at the time of diagnosis, distant site of 1st recurrence, 1st non-prostate recurrence and metastatic lymph nodes). Review revealed that these data elements were being poorly collected in quantity (<1% of cases with a valid response) and quality (75% cases with discrepancies when compared to clinical staging or PSA recurrence data). A discussion with the cancer registrars revealed inconsistent application of the definitions for distant metastasis and recurrence. Consequently, the CDEs were modified and re-defined, and subsequent re-evaluation of the collected cases revealed improved data collection.
The first generation CDEs that were determined to result in poorly collected data from all the sites were eliminated and no longer collected after a review of the initial 2600 cases entered into the Resource. For example, a CDE for smoking history attempted to collect multiple values such as "current smoker, never smoked, past smoker, smoker (current or past unknown), and unknown". These values were available for only a limited number of cases (44% of cases with data), which many of them were inconsistently applied because of ambiguous definitions and subjective reporting from the clinical charts. These concerns led the CPCTR Coordinating Committee [18] to eliminate this CDE. However, each institution that elected to resume collecting discontinued CDEs were allowed to locally, but were not reported to the central database.
The Coordinating Committee also re-evaluates the resource as a whole periodically to meet the needs of the prostate research community. The CDE sub-committee is charged to add new CDEs based on the needs of the researchers and addition of new resource materials (e.g., biopsy only specimens, frozen matrix blocks, TMAs, and metastatic tissue blocks). The Coordinating Committee has added dates to many of the CDE categories to be able to examine the timeline of major events that may occur for each patient during his course of disease. This "event table" has resulted in a better picture and assessment of the inter-related characteristics of patient treatments and outcomes.
Once the changes of the CDEs were approved, the local and central databases were modified accordingly. The existed data were programmed to be mapped and stored to the updated databases. The local dataset from each site was queried and sent to the central database following the requested format and validated to ensure the correction of the CDEs updating.
Discussion
In order to develop any biospecimen resource with high quality specimen annotation, the initial process of building the resource involves significant time and commitment from many experts from various disciplines. Open discussions and input from all potential parties with a stake in the outcome is crucial to any such developmental work. The process of developing the CDEs for the CPCTR has attested that this approach can successfully lead to the implementation of robust prostate tissue CDEs that guide the collection of quality data at over 18 different institutions or hospitals [6,19,20].
Success depends on the ability to collect data using CDEs that have been evaluated by a working group that provides inputs from various experts. The CPCTR CDE sub-committee included organ specific clinicians (pathologists and urologists), informaticians, biostatisticians, data managers, cancer registrars, and research scientists. Clinicians were primarily responsible for providing the foundation of data elements as they reflected the current standard of information used in patient care decisions, while attempting to project at least five years into the future for additional data that may become clinically significant. Likewise, research scientists provided input on data elements that would be crucial in the evaluation of current or proposed prostate cancer research with respect to the detection, diagnosis, prognostication, and treatment of prostate cancer. Thus the result was the creation of datasets that should provide value to the research community when requested from the resource for years to come.
Local data collection methods vary at each institution based on personnel. Some sites have cancer registries responsible for obtaining patient follow up data, while others obtain data from their registry systems or have independent nurses or data managers who extract data from Urology offices by reviewing charts. Thus, it was important to include these nurses, data managers, and cancer registrars who are the main data collectors for the tissue banking resource in CDE development. Their input on the types of data and meta-data available for collection proved to be crucial in aggregating highly quality annotation data for the bio-specimens. Moreover, the definitions of the CDEs and their associated metadata need to be clearly understandable to all those who collect data. For example, in order to collect quality data, the collectors need to understand 1) the fundamental definition of the data element (i.e., date of diagnosis), 2) how that data element will be collected (e.g. 11/2003 vs. Nov. 2003 vs. 11/03, etc), 3) what are the consensus acceptable values or codes are for the data element (e.g., precise date of birth, not calculated from clinical records where the "patient appears to be a well developed 75 year old"), and 4) what the acceptable data format is for inclusion into the central database (e.g., dates as integers not character strings). Through the use of ISO 11179 compliance standards, the goals of collecting annotation data of high quality was achievable, and emphasize the consensus approach used by CPCTR as being critical to successful CDE creation. Although the concept of formalized metadata is fairly straight forward, it has been rarely incorporated by clinical and research groups building databases [14].
Another demonstration of the benefit by CPCTR CDEs is evident from the implementation of the TMA data exchange specification sponsored by the Association of Pathology Informatics (API) [21]. This specification has been used to provide a supplemental XML (Extensible Markup Language) file of the data describing each of the cores in the TMA slides provided to researchers through the CPCTR [22]. The CDEs allow the Resource to directly port data elements and associated metadata directly into a TMA file that complies with the API's TMA specification document and that contains a protected namespace for the CPCTR metadata. These study cases show the examples that well developed CDEs can benefit comparative research and in-depth analysis of data among multiple institutions and studies. The only way in which information from multiple databases can truly be shared and made useful is through the careful use of clearly defined metadata and CDEs [23].
Informaticians and database developers provided the structural link that brought the CDEs together in the database, addressed technical issues, and provided guidance related to implementation of the CDEs at local institutions. The success of the CPCTR CDEs is shown by their implementation at four separate institutions using four different databases, thus demonstrating the ease at which these standards can be copied and distributed across institutions. Regardless of which type of database is used locally, each site is responsible for mapping their data dictionary to the CPCTR CDEs when their data is sent to IMS to be shared. At each individual institution additional data is collected that is pertinent to institution-specific research goals. Yet this incorporation of the common CDEs allows institutions to share data and results across groups while maintaining the autonomy of their research objectives.
Furthermore, having the ability to collect high quality "simple" data elements that have been agreed upon by a working group is crucial for the overall quality of quantitative analysis of inter-institutional data. Collecting simple, yet uniform and comprehensive data annotations in a common database for research across multiple institutions, each with various capabilities of collecting the data (manual review of medical charts, cancer registry systems, and interfaces to legacy systems), vastly increases the statistical power of research efforts and has the potential to identify common trends and issues in cancer care. It is critical that these trends and issues be addressed if we are to find methods to reduce the cancer burden and cancer pain and suffering as is the goal of the NCI [24]. The value of tissue banks and the informatics that support these goals are clearly outlined in the NIH and NCI strategic roadmaps [25].
Conclusion
Recently, there has been an increasing number of international [26-32] as well as national and state-wide [20,33,34] initiatives that have promoted formation of large research consortia and encourage these groups to share both tissue and data. Currently, many tissue banks such as the CBCTR [3,4], CHTN [35], CFR [36], SPOREs [37], EDRN [13,38] and the PCABC [20] involve multiple institutions. These biorepositories vary in their data collection and tissue collection methodologies. However, the necessity for well annotated tissues that can be re-annotated with experimental data has driven many of these multi-institutional collaborations to develop standards of sharing data with other groups. Currently, the CPCTR CDEs are specifically related to the available prostate tissue resources and clinical data, while experimental data generated from these tissue specimens are not required to be submitted to the resource. However, publications resulting from the use of CPCTR tissues are obliged to credit the Resource, allowing the results to be correlated to or compared with other studies using similar CDE standards. Other initiatives such as the Shared Pathology Informatics Network or SPIN [19], the Early Detection Research Network or EDRN [38] and the Cancer Biomedical Informatics Grid initiative or CaBIG [39] can perform follow-up studies by linking their results to CPCTR derived studies by using the common CPCTR CDEs. This also allows for meta-analysis of data across studies through the CPCTR CDEs, resulting in improved statistical power and further detailed analysis. Thus, expanding the CPCTR dataset by combining tissue with experimental data will have tremendous value in enhancing cancer research [40].
Based on the experience of developing CDEs for the CPCTR, the following sequential strategies can be recommended for other research groups involved in future CDE development efforts.
Initial several months to a year:
• Decide what CDEs the resource will need using a committee driven consensus process that include all major stakeholders
• Utilize as a starting point similar CDE initiatives already developed by others and build upon their standards
• Consult a variety of experts, including those that will be collecting the data particularly tissue bankers, cancer registrars and data managers
Next few months:
• Draft a CDE data dictionary which includes not only the structured data, but also precise data field definitions and a consideration of metadata (data that describes the original data)
• Identify the essential/required data elements and ratify them through a consensus process
• Modify or approve CDEs after discussions with all key parties and build consensus among them and any other external experts
• Create corresponding data entry paper forms/data entry interface to central database
Subsequent few months:
• Implement CDEs
• Test/Pilot phase: sharing of data with central database
Continuously ongoing efforts:
• Re-evaluate CDEs and their data values (every year or after set accrual targets)
• Develop quality assurance, quality control and quality improvement protocols to fully develop the CDEs (minimum of once per year quarterly of semi annually or 10% of new data set, which is the current norm for CPCTR)
Abbreviations
AJCC – American Joint Committee on Cancer
API – Association for Pathology Informatics
CaBIG – Cancer Biomedical Informatics Grid
CAP – College of American Pathologist
CBCTR – Cooperative Beast Cancer Tissue Resource
CDE – Common data elements
CDP – Cancer Diagnosis Program
CFR – Cancer Family Registries
CHTN – Cooperative Human Tissue Network
CPCTR – Cooperative Prostate Cancer Tissue Resource
EDRN – Early Detection Research Network
HGPIN – High-grade prostatic intraepithelial neoplasia
HIPAA – Health Insurance Portability and Accountability Act
HTTP – HyperText Transfer Protocol
ISO – International Organization for Standards
IRB – Institutional Review Board
LN – Lymph nodes
Mets – Metastasis
NAACCR – North American Association of Central Cancer Registries
NCI – National Cancer Institute
NIH – National Institutes of Health
PCABC – Pennsylvania Cancer Alliance Bioinformatics Consortium
PI – Principle Investigator
PSA – Prostate Specific Antigen
QA – Quality assurance
RAND – Rand Corporation, Inc.
REP – Research Evaluation Panel
RFA – Request for Applications
SEER – Surveillance, Epidemiology, and End Results
SPIN – Shared Pathology Informatics Network
SPOREs – Specialized Programs of Research Excellence
TMA – Tissue Microarray
XML – Extensible Markup Language
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AAP was the chair for the data manager's sub-committee and wrote the first draft of the manuscript. MJB, who was the chair for the CDE sub-committee and the Coordinating Committee, was responsible for leading the efforts of developing the CDEs. AKB, JJB, MB, MWD, RD, JG, JM, JO, and KFT assisted in the development of the CDEs and incorporation of other existing standards. All authors reviewed and commented on successive drafts of the manuscript and have provided the first author with approval of the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional file 1
CPCTR CDE dictionary CPCTR common data elements' data dictionary with full description and meta-data for each of the CDEs.
Click here for file
Additional file 2
CPCTR CDE paper forms. CPCTR paper forms used for data collection of CDEs by data managers. These forms are used for collecting data at remote locations (i.e. physician offices) and then entered into the database.
Click here for file
Acknowledgements
Supported in part by NIH Grants from the National Cancer Institute U01 CA86772 and P30 CA13343 (NYU School of Medicine), U01 CA86735 (University of Pittsburgh), U01 CA86743 (Medical College of Wisconsin and University of Illinois-Chicago), and U01 CA86739 (George Washington University).
We acknowledge all the following contributors to the development of the resource (in alphabetical order by last name):
George Washington University: John Bayerl, Farrell Hartigan, Judith Horn, Charmaine Mckenzie, and Deana McRae.
MCW: Trisha St. Laurence-Urbaniak, Pat Recely, Martha See, William See, Peter Langenstroer, Robert Donnel, Hongyung Choi, Jeanne Hryciuk, Douglas Chausow, Michael Whittaker, Mary Fernandez, Lawrence Clowry, Rueben Eisenstein, Andrea Samaniego, Tracey Brodzeller, Andrea Kahler, Sharon Fuller, and Sushma Kaul.
New York University: Richard Buchsbaum, Soraya Kernizan, Xiangtian Kong, Marti Ksionsk, Kyle Kuhn, Payal Patel, Joanne Schmoll, Lindsay Stanton, and Hiroko Watanabe.
University of Pittsburgh: Marie Acquafondata, Sheldon Bastacky, Michelle Bisceglia, Vicky Chu, Dilip Gupta, Drazen Jukic, Marianne Notaro, Jennifer Steudler, Susan Urda, and Tracy Wagner (funded members). We also would like to specially recognize the unfunded members of the Pitt team including Wendy Hillard and Alena Sikorova.
Figures and Tables
Figure 1 CPCTR Organization of the Resource. The Coordinating Committee determines the types of biospecimens the CPCTR will provide. The Research Evaluation Panel (REP) from NCI is the committee in consultation with the Coordinating group. The sub-committees, pathology, CDE, data manager coordinate each other to develop the CDEs for different types of biospecimens that CPCTR will collect.
Figure 2 CPCTR metadata dictionary application. This is a screen shot of the Oracle mid-tier application used for adding or modifying active CDEs along with the associated metadata. It is also used for generating the CPCTR public query tool available at
Figure 3 CPCTR CDE categories. Four main categories of CDEs and the sub-data types collected. Detail description of each CDE under these four main categories and its sub-groups can be found in the CPCTR CDE data dictionary.
Figure 4 CPCTR Critical Data Fields. The critical data fields are divided into two categories$: 1) Required fields, which are essential and must be entered into the database for a case to be accepted. 2) Conditional required fields, which must be filled out when the respective tissue matrix is entered into the Resource.
Figure 5 Re-Evaluation of CDEs. The flow chart describes the re-evaluation process involved for validating the CDEs. Any changes made after this process, usually eliminates discrepancies and difficulties with the data collection process. In addition, the database also shows an increase in number of fields being populated with valid values and a decrease in "unknown" values.
==== Refs
Eiseman E Rand Corporation Case studies of existing human tissue repositories: "Best practices" for a biospecimen resource for the genomic and proteomic era 2003 Santa Monica, CA: RAND
Cooperative Prostate Cancer Tissue Resource: Release date April 29, 1999, RFA CA-99-012, National Cancer Institute
The Cooperative Breast Cancer Tissue Resource
Glass AG Donis-Keller H Mies C Russo J Zehnbauer B Taube S Aamodt R The Cooperative Breast Cancer Tissue Resource: Archival tissue for the investigation of tumor markers Clin Cancer Res 2001 7 1843 1849 11448894
Cooperative Prostate Cancer Tissue Resource
Melamed J Datta M Becich M Orenstein J Dhir R Silver S Fidélia-Lambert M Kadjacsy-Balla A Macias V Patel A Walden P Bosland M Berman J the Cooperative Prostate Cancer Tissue Resource The Cooperative Prostate Cancer Tissue Resource: A specimen and data resource for cancer researchers Clin Cancer Res 10 4614 4621 2004 Jul 15 15269132
Department of Health and Human Services. 45 CFR (Code of Federal Regulations), 164.514(6)(2)(i). Standards for Privacy of Individually Identifiable Health Information (final)
The Cooperative Prostate Cancer Tissue Resource Manual of Operations
Fleming ID American Joint Committee on Cancer, American Cancer Society, American College of Surgeons AJCC Cancer Staging Manual 1997 5 Philadelphia: Lippincott-Raven
NAACCR Data Standards for Cancer Registries
The CAP cancer protocols: Checklist for prostate gland
The NCICB's Cancer Data Standards Repository (caDSR)
Winget MD Baron JA Spitz MR Brenner DE Warzel D Kincaid H Thornquist M Feng Z Development of common data elements: The experience of and recommendations from the early detection research network Int J Med Inform 2003 70 41 48 12706181 10.1016/S1386-5056(03)00005-4
Solbrig HR Metadata and the reintegration of clinical information: ISO 11179 MD Comput 2000 17 25 28 10842979
CPCTR CDE documentation
Information Management Services, Inc
CPCTR public database
CPCTR Coordinating Committee
The Shared Pathology Information Network website
Pennsylvania Cancer Alliance Bioinformatics Consortium website
Berman JJ Edgerton ME Friedman BA The tissue microarray data exchange specification: A community-based, open source tool for sharing tissue microarray data BMC Med Inform Decis Mak 2003 3 5 12769826 10.1186/1472-6947-3-5
Berman JJ Datta M Kajdacsy-Balla A Melamed J Orenstein J Dobbin K Patel A Dhir R Becich MJ The tissue microarray data exchange specification: Implementation by the Cooperative Prostate Cancer Tissue Resource BMC Bioinformatics 2004 5 19 15040818 10.1186/1471-2105-5-19
Berman JJ Pathology data integration with extensible markup language Hum Pathol 2005 36 139 145 15754290 10.1016/j.humpath.2004.10.013
The National Advanced Technologies Initiative for Cancer
NIH Roadmap: Accelerating medical discovery to improve health
ATIP03.042: The Biobank Japan Project
Australian Prostate Cancer Bio-Resource Website
McCaffrey P Iceland's database tussle 2003 CIO Magazine, April 1
Triendl R Japan launches controversial Biobank project Nat Med 2003 9 982 12894146 10.1038/nm0803-982b
Swedish National Biobank Program website
The UK Biobank Project website
Spinney L UK launches tumor bank to match maligned Biobank Nat Med 2003 9 491 12724749 10.1038/nm0503-491a
The Georgia Cancer Coalition website
The Kentucky Lung Cancer Research Program website
The NCI's Cooperative Human Tissue Network (CHTN)
The NCI's Cancer Family Registries (CFR)
The NCI's Specialized Programs of Research Excellence (SPOREs)
The NCI's Early Detection Research Network website (EDRN)
The NCI's Cancer Biomedical Informatics Grid (caBIG) website
The National Biospecimen Network (NBN) Blueprint
|
16111498
|
PMC1236914
|
CC BY
|
2021-01-04 16:03:06
|
no
|
BMC Cancer. 2005 Aug 21; 5:108
|
utf-8
|
BMC Cancer
| 2,005 |
10.1186/1471-2407-5-108
|
oa_comm
|
==== Front
BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1111614304010.1186/1471-2407-5-111Research ArticleCan p503s, p504s and p510s gene expression in peripheral-blood be useful as a marker of prostatic cancer? Cardillo Maria Rosaria [email protected] Vincenzo [email protected] Antonio [email protected] Laura [email protected] Stefano [email protected] Silverio Franco [email protected] Department of Experimental Medicine and Pathology (Section of Pathologic Anatomy). University "La Sapienza", Viale Regina Elena, 161 Rome, Italy2 Department of Urology "U. Bracci", University "La Sapienza", Viale del Policlinico, 156 Rome, Italy3 Department of Surgical Science, University "La Sapienza" Viale Regina Elena, 161 Rome, Italy2005 5 9 2005 5 111 111 28 12 2004 5 9 2005 Copyright © 2005 Cardillo 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 aim of the study was to investigate whether p503S, p504S and p510S gene expression in peripheral-blood be useful as a diagnostic or prognostic marker of prostatic cancer.
Methods
Circulating cells were identified by reverse transcription-polymerase chain reaction (RT-PCR) to detect p503S, p504S and p510S mRNA in peripheral blood (PB) from 11 patients with treated prostatic carcinoma (CaP), 11 with newly-diagnosed untreated CaP and 20 with benign prostatic hyperplasia (BPH) (controls).
Results
RT-PCR amplified P503S in 7 of 11 untreated and 2 of 11 treated patients with CaP and 5 of 20 with BPH; p504S in 7 of 11 untreated and in 9 of 11 treated patients with CaP and 11 of 20 with BPH; whereas it amplified p510S in all subjects with CaP and in 15 of 20 with BPH.
Conclusion
These findings suggest that the investigated genes are poorly specific and probably of little use as diagnostic or prognostic prostatic markers in peripheral blood for monitoring disease progression and recurrence.
==== Body
Background
Microarray approaches have identified three prostate tissue and cancer-prostate-specific genes: p503S, a 241-amino acid protein that encodes human tetraspan NET-1, a member of the tetraspanin/TM4SF family, involved in cancer metastasis) [1], P504S, also referred to as the AMACR gene, encodes human α-methylacyl-CoA racemase, a 382-amino acid protein involved in the conversion of R-stereoisomers of branched-chain fatty acids to S-stereoisomers [2-6] and p510S, identified as the human ABC transporter MOAT-B [7]. All these genes are overexpressed in prostate tumor or normal prostate tissue or both and are considered clinical biomarkers of prostate cancer (CaP)[8]. The immune response against AMACR has been used as a serum biomarker for CaP [9,10] and the quantification of AMACR transcripts in prostatic secretions shown to be predictive of CaP [11].
Attempts to detect AMACR in circulation have been disappointing and no molecular studies have yet sought p503S and p510S circulating cells by reverse transcriptase-polymerase chain reaction (RT-PCR) assay of peripheral blood (PB) from patients with CaP.
The aim of this study was to investigate whether p503S, p504S and p510S genes are diagnostic or prognostic prostate-specific markers for monitoring disease progression and recurrence. Extending our previous research investigating biomarkers that are clinically important to distinguish patients with prostate cancer at risk for early relapse from patients in clinical remission [12-14], in this study, we analyzed by RT-PCR assays p503S, p504S and p510S gene markers in PB from 42 patients with treated or untreated CaP and with benign prostatic hyperplasia (BPH) (controls).
Methods
Patients and control tissue specimens
The 42 participants included in this study were consecutively selected from patients treated at the Department of Urology "U. Bracci", University of Rome "La Sapienza". We studied three clinical groups. The first group comprised 11 patients with CaP (mean age at diagnosis 68.9 years; range 45–89) who underwent androgen ablation (AA) followed by retropubic radical prostatectomy (RP) and bilateral lymphadenectomy between January 1997 and December 1999. Tumors were graded using the Gleason system for histologic grading of CaP: two tumors were Gleason score sum 4, 7 Gleason sum 7 and 2 Gleason sum 8. Tumors were then evaluated clinico-pathologically using the tumor-node metastatic (TNM) staging system 2002: 7 tumors were organ-confined diseases (pT2) and 4 were extra-prostatic diseases (pT3). Follow-up after treatment ranged from 3–84 months (mean, 27 months). At the time of venipuncture for PB, we distinguished two subgroups: patients who responded to the first 24 months of treatment without clinical progression; and AA resistant patients who received neoadjuvant androgen deprivation therapy because recurrent disease developed within 24 months after RP. We defined recurrence by serum PSA levels, transrectal ultrasound (TRUS) and computed tomographic (CT) scan. Men were considered disease free after RP if they had serum PSA levels less than 0.1 ng/ml. The second subgroup comprised 11 patients with elevated serum PSA levels (more than 4.0 ng/ml), who had received a diagnosis of CaP based on TRUS-guided-prostate biopsy (1 tumor was Gleason sum 6 and 10 were Gleason sum 7). All 11 patients had organ-confined disease: 4 were pT1 and 7, pT2. In patients treated for CaP, PB samples were obtained 27 to 84 months (mean 55 months) after treatment ended; in patients who had received a diagnosis of CaP based on clinical and biochemical evidence of malignancy, PB samples were obtained before biopsy and before treatment began. The third group (control group) comprised 20 men 50 years of age or older (mean age 58.7 years; range 45–79) with no known malignancy, with a histologic diagnosis of BPH in tissue obtained by fine-needle biopsy of the prostatic glands and on conventional clinical data (serum PSA levels less than 4.0 ng/ml, and digital rectal examination and TRUS negative for CaP). The control group also included a PB sample from a patient who underwent radical cystectomy for bladder cancer (G3T1) and 6 neoplastic and non neoplastic prostatic tissues; 6 tissues from various organs (adrenal gland, heart, myometrium and uterine cervix); and 6 prostatic (SV48, LS147D, SCOV 3 and MCF7) and non prostatic cell lines (PC3 and LNCaP).
Reverse-transcription polymerase-chain-reaction (RT-PCR)
Total RNA was extracted from whole blood cells and cell lines using the technique previously described [12-15]. Total RNA from frozen tissue was extracted using the protocol developed by Ambion, and detailed in the RNA isolation Kit (Ambion Inc., U.S.A). cDNA (250 ng) from each sample was amplified by PCR using the following primers: sense, 5'TGCCCTCGTGACGTTCTTCT3' and antisense 5'TCTTTCTTGATGGCAGGCACTAC3' of P503, 136 bp (GeneBank accession number AF065388 (16) using 35 cycles (94°C for 30", 60°C for 30" and 72°C for 60 sec); sense 5'AAATGGTTATCATTAGGGCTTTTGA3' and antisense 5'TTCCTTTTTCACTAGAACCCATTCA3' of P504S, 149 bp (GeneBank accession number 4204097 [17] using 35 cycles (94°C for 30", 55°C for 30" and 72°C for 60 sec); sense 5'TTGAACAGCTACTACGGTCAATGTATT3'and antisense 5' GCAGAGAGCAACCGATGTTTT3' of P510S, 96 bp (GeneBank accession number AF071202 [18] using 35 cycles (94°C for 30", 57°C for 30" and 72°C for 60 sec) by Platinum Taq DNA polymerase (Gibco BRL), according to the manufacturer's protocols. The integrity of RNA was checked in a preliminary PCR reaction for a human β-actin 743 base-pair (bp) fragment [19]. Each RT-PCR experiment included a non retrotranscribed sample as a negative control and c-DNA from LNCaP cell lines as a positive control. PCR-amplified P510S sequence specificity was checked by direct sequencing of the amplified product by the automated DNA sequencer A.L.F. (Pharmacia, Freiburg, Germany). The sequencing reaction was run on an Abby Applied 377 Sequencer (Applied Biosystems). Band identity was verified by comparison with the genomic and cDNA P510S sequences at the GeneBank using the BLAST program.
Results
RT- PCR analysis amplified p5103S-, p504S- and p510S-mRNA in PB from nearly all patients who underwent AA and RP, who had recurrent disease and p503S and p504S in only 2 (40%) of the 5 patients without recurrence (Fig. 1; Table 1). Comparing the RT-PCR findings with the histological Gleason scores we found p503S- and p504S-positive cells in a higher percentage of patients with higher-grade Gleason scores than lower-grade scores. P503S were more often expressed in extraprostatic disease (pT3N1 and pT3N2) and p504S in organ-confined disease (pT2N0). The p510S gene was amplified in all patients regardless of the clinical subgroup (Table 2).
Two samples from 11 untreated patients with biopsy-proven CaP contained p503S mRNA. This gene was expressed significantly less frequently than p504S and p510S (18% vs 81 and 100% (P <.01, by chi-square test) (Fig. 2; Table 1). PB from the control patient with non prostate malignancies, who underwent radical cystectomy for transitional-cell carcinoma (G3T1), contained p504S- and p510S-but no p503S-positive cells (Fig. 2, lane 12). Nearly all untreated patients with Gleason score 6–7 tumors and organ-confined disease had p504S- and p510S-positive cells, whereas only 2 of the 11 patients had p503S-positive cells. p503S-positive cells were detectable in patients with stage pT2 and undetectable in those with stage pT1 disease. p510S-positive cells were expressed in similar percentages of patients with pT1 and pT2 disease (Table 2).
In PB samples from the control group (BPH), PCR amplified more p504S- and p510S- than p503S-positive cells (55 and 75% vs 25%) (Fig. 3). Nearly all the prostatic and non prostatic-tissues and cell lines used as controls expressed p503S-, p504S-, and p510S-mRNA (Table 1) (Fig. 4). DNA sequencing of the amplified 96 bp fragment confirmed the specificity of these primers for the p510S gene (Fig. 5).
Discussion
These findings argue against the usefulness of the p503S, p504S and p510S genes as diagnostic and probably against their usefulness as prognostic prostate-specific markers for monitoring disease progression and recurrence in PB from patients with CaP. Our findings obtained by RT-PCR assay contrast with Zelie et al [11] et al who found the RT-PCR-AMACR test non-invasive and useful in predicting the presence of CaP in prostatic secretions. When we compared our findings for p503S, p504S and p510S with our earlier study investigating prostatic specific antigen (PSA) [3], as potential markers for prostate tumors, all these three genes were more frequently expressed than PSA in PB from patients with treated CaP (63%, 63% and 100% vs 18%) and untreated CaP (18%, 81% and 100% vs 9%) and in PB from BPH (25%, 55% and 75% vs 0). In benign and malignant prostate tissue, p510S was more strongly expressed than PSA, or p503S and p504S (100% vs 66% vs 66%), whereas in prostatic cell lines, PSA and p503S, p504S and p510S were similarly expressed (100%). In non prostatic tissue and cell lines, PSA was not expressed; in contrast, in non prostatic samples and cell lines, p503S was more expressed than p510S and p504S (100% vs 33% vs 16%) and cell lines (100% vs 50% vs 50%). Hence, the increased expression of PSA mRNA in prostate tissue suggests that PSA is more specific than p503S, p504S and p510S as a marker of cancer progression and dissemination.
The ubiquitous presence of p503S, p504S and p510S in PB of patients with BPH, in non prostatic tissue and cell lines therefore argues against their potential value as biomarkers of prostate disease. Compared with PSA [13], none of the three newly-identified putative CaP markers we studied were sufficiently sensitive or specific to distinguish patients in complete clinical remission from patients with occult metastatic disease at high-risk for relapse. The numerous false-positive findings in our assay and the reported presence of p503S, p504S and p510S in normal hematopoietic tissues also from patients without prostate malignancies [20], could depend on illegitimate expression of gene transcripts. This question awaits an answer from studies designed to investigate the whole p503S, p504S and p510S gene sequence. For RT-PCR analysis of PB cells in this study we used primers for three new putative genes (p503S, p504S, and p510S) whose prostate cancer specificity has already been confirmed by Northern blot, real-time PCR and immunohistochemical assay on prostatic tissue [3-8]. Although AMACR may be potentially useful as a tissue biomarker for prostate cancer [3,4], the limitations of this assay make it impossible to use an autoimmune response against AMACR as a means of detecting this antigen in the serum [10]. Nor did a novel approach to predict the presence of CaP from prostatic secretions prove predictive of CaP [9]. Whether these new markers have potential clinical utility as sensitive and specific indicators of CaP progression therefore warrants further research using quantitative RT-PCR technique to study the entire gene sequences.
Conclusion
We provide the first evidence of the ubiquitous presence of p503S, p504S and p510S in PB of patients with BPH, in non prostatic tissue and cell lines. These findings suggest that the investigated genes are poorly specific and probably of little use as diagnostic or prognostic prostatic markers in peripheral blood for monitoring disease progression and recurrence.
List of abbreviations used
RT-PCR – reverse transcription-polymerase chain reaction
CaP – prostatic carcinoma
BPH – benign prostatic hyperplasia
AMACR – α-methylacyl-CoA racemase
PB – peripheral blood
AA – androgen ablation
RP – radical prostatectomy
TRUS – transrectal ultrasound
CT – computed tomographic
TNM – tumor-node metastatic
PSA – prostatic specific antigen
Competing interests
The author(s) declare that they have no competing interests
Authors' contributions
MRC provided input into the design of these studies, assistance with development of the assays used for these molecular genetic studies in the laboratory of Uropathology at Department of Experimental Medicine, Rome "La Sapienza" University, data analysis and writing of the manuscript.
AC provide assistance with patients diagnoses undergone TRUS-guided-prostate biopsy and collection of specimens and clinical date.
VG provide assistance with patients diagnoses and collection of specimens and clinical data, and supervised all aspect of the work performed for this paper in the Department of Urology, Rome "La Sapienza" University.
LG provided collection of specimens and clinical data, and assisted with data analysis.
SM provided assistance with patient diagnoses and with data analysis.
FDS provided assistence with patients undergone surgery, collection of specimens and clinical data, and supervised all aspect of the work performed for this paper in the Department of Urology, Rome "La Sapienza" University.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study was supported by grants from the Ministero Università della Ricerca Scientifica e Tecnologica (M.U.R.S.T) of the Facoltà and Ateneo 2002–2003, and the Associazione Ricerca di Base in Urologia (A.R.B.U).
Figures and Tables
Figure 1 Analysis of gene expression levels of p503S, p504S and p510S by RT-PCR using β-actin yielded a 740 base-pair band as internal control, in whole blood from 11 patients after radical prostatectomy and androgen ablation (lanes 1 to 11); negative control, no RNA (lane 12); positive control (LNCaP cell line) (lane13); MW (molecular weight).
Figure 2 Analysis of gene expression levels of p503S, p504S and p510S by RT-PCR using β-actin yielded a 740 base-pair band as internal control, in whole blood from 11 patients with CaP after biopsy (lanes 1 to 11) and one patient with transitional cell carcinoma of the bladder (lane 12) as control. Negative control (no RNA) (lane 13); positive control (LNCaP cell line) (lane 14); MW (molecular weight).
Figure 3 Analysis of gene expression levels of p503S, p504S and p510S by RT-PCR using β-actin yielded a 740 base-pair band as internal control in whole blood from 17 of 20 patients with benign prostatic disease (BPH).
Figure 4 Analysis of gene expression levels of p503S, p504S and p510S by RT-PCR using β-actin yielded a 740 base-pair band as internal control in frozen benign (lanes 1, 3, and 6) and neoplastic (lanes 2, 4, and 5) prostatic tissue; non prostatic tissue (adrenal gland (lane 7); heart (lanes 8 to 10); myometrium (lane 11); uterine cervix (lane 12) and human cell lines: SV48 (lane 13); LS147D (lane 14); SCOV 3 (lane 15); MCF7 (lane 16); PC3 (lane 17); and LNCaP (lane 18). Negative control no RNA (lane 19); MW (molecular weight).
Figure 5 Nucleotide sequence of P510S cDNA.
Table 1 RT-PCR assay in peripheral blood, in different tissues and cell lines.
SAMPLES RT-PCR P503S P504S P510S
PB No (%) No (%) No (%)
group 1
AA and RP (n = 11) positive 7 (63.63) 7 (63.63) 11 (100)
negative 4(36.36) 4(36.36) 0
group 2
Pca in prostate biopsy (n = 11) positive 2 (18.18) 9 (81.81) 11 (100)
negative 9 (81.81) 2 (18.18) 0
group 3
BPH in prostate biopsy (n = 20) positive 5 (25) 11 (55) 15 (75)
negative 15 (75) 9 (45) 5 (25)
Frozen tissues
Neoplastic and non neoplastic positive 4 (66.66) 4 (66.66) 5 (83.83)
prostate tissues (n = 6) negative 2 (33.33) 2 (33.33) 1 (16.16)
Non prostatic tissues (n = 6) positive 6 (100) 1(16.66) 2 (33,33)
negative 0 5 (83.33) 4 (66.66)
Cell lines (n = 6)
prostatic (n = 2) positive 2 (100) 2 (100) 2 (100)
negative 0 0 0
non prostatic (n = 4) positive 4 (100) 2 (50) 2 (50)
negative 0 2 (50) 2 (50)
PB = peripheral blood
RT = radical prostatectomy
AA = androgen ablation
Table 2 p503S, p504S and p510S positive cells in peripheral blood of patients with prostate cancer (CaP) in relation to Gleason score and TNM stage.
P503S+ P504S+ P510S+
Group 1 (treated CaP) No No No
Total (n = 11) patients (%) patients (%) patients (%)
subtype 1 without recurrence (n = 5) 2 (40) 2 (40) 5 (100)
subtype 2 with recurrence (n = 6) 5 (83) 5 (83) 6 (100)
Gleason score
4–7 (n = 9) 4 (44.44) 6 (66.66) 9 (100)
8 (n = 2) 2 (100) 2 (100) 2 (100)
TNM
Organ-confined disease (n = 7)
T2 N0 M0 (n = 7) 4 (57.14) 5 (71.42) 7 (100)
Extraprostatic disease (n = 4)
T3 N0 M0 (n = 2) 1 (50) 1 (50) 2 (100)
T3 N1M0 and T3 N2M0 (n = 2) 2 (100) 1 (50) 2 (100)
Group 2 (untreated CaP)
Total (n = 11)
Gleason score
6–7 (n = 11) 2 (18.18) 9 (81.81) 11 (100)
TNM
Organ-confined disease (n = 11)
T1 (n = 4) 0 3 (75 4 (100)
T2 (n = 7) 2 (28.6) 6 (85.7) 7 (100)
==== Refs
Claas C Seiter S Claas A Savelyeva L Schwab M Zoller M Association between the rat homologue of CO-029, a metastasis-associated tetraspin molecule and consumption coagulopathy J Cell Biol 1998 141 267 280 9531564 10.1083/jcb.141.1.267
Reichel C Brugger R Bang H Geisslinger G Brune K Molecular cloning and expression of a 2-arylpropionyl-coenzyme A epimerase; a key enzyme in the inversion metabolism of ibuprofen Mol Pharmacol 1997 51 576 582 9106621
Rubin MA Zhou M Dhanasekaran SM Varambally S Barrette TR Sanda MG Pienta KJ Ghosh D Chinnaiyan AM α-Methylacyl Coenzyme A Racemase as a tissue biomarker for prostate cancer JAMA 2002 287 1662 1670 11926890 10.1001/jama.287.13.1662
Luo J Zha S Gage WR Dunn TA Hicks JL Bennett CJ Ewing CM Platz EA Ferdinandusse S Wanders RJ Trent JM Isaacs WB De Marzo AM α-Methylacyl CoA Racemase: a new marker for prostate cancer Cancer Research 2002 62 2220 2226 11956072
Kuefer R Varambally S Zhou M Lucas PC Loeffler M Wolter H Mattfeldt T Hautmann RE Dunn RL Chinnaiyan AM Rubin MA α-Methylacyl CoA Racemase: expression levels of this novel cancer biomarker depend on tumor differentiation Am J Pathol 2002 161 841 848 12213712
Zheng SL Chang B-lI Faith DA Johnson JR Isaacs SD Hawkins GA Turner A Wiley KE Bleecker ER Walsh PC Meyers DA Isaacs WB Xu J Sequence variants of α-Methylacyl CoA Racemase are associated with prostate cancer risk Cancer Research 2002 62 6485 6488 12438241
Xu J Stolk JA Zhang X Silva SJ Houghton RL Matsumura M Vedvick TS Badaro R Reed SG Identification of differentially expressed genes in human prostate cancer using subtraction and microarray Cancer Research 2000 60 1677 1682 10749139
Horoszewicz JS Kawinski E Murphy GP Monoclonal antibodies to a new antigenic marker in epithelial prostatic cells and serum of prostatic cancer patients Anticancer Res 1987 7 927 935 2449118
Carter HB Isaacs WB Improved Biomarkers for prostate cancer: A definite Need JNCI 2004 96 813 815 15173257
Sreekumar A Laxman B Rhodes DR Bhagavathula S Harwood J Giacherio D Ghosh D Sanda MG Rubin MA Chinnaiyan AM Humoral Immune response to α-methylacyl-CoA racemase and prostate cancer JNCI 2004 96 834 843 15173267
Zielie PJ Mobley JA Ebb RG Jang Z Blute RD Ho SM A novel diagnostic test for prostate cancer emerges from the determination of alpha-methylacyl-coenzyme a racemase in prostatic secretions J Urol 2004 172 1130 1132 15311056 10.1097/01.ju.0000133560.87118.4d
Gandini O Luci L Stigliano A Lucera R Di Silverio F Toscano V Cardillo MR Is DD3 a new prostate-specific gene? Anticancer Research 2003 23 305 308 12680228
Cardillo MR Gentile V Di Silverio F Tumor target prostate specific membrane antigen (PSMA) and its regulation in prostate cancer J Cellular Biochemistry 2004 93 641 643 10.1002/jcb.20244
Cardillo MR Di Siverio F Gentile V The reliability of PSA circulating cells as markers of metastatic prostate cancer Clin Lab Invest 2004 64 687 690 10.1080/00365510410003101
Chomczynski P Sacchi N Single step methods of RNA extraction by acid-guanidinium-thiocyanate-phenol-chloroform extraction Anal Biochem 1987 162 156 159 2440339 10.1016/0003-2697(87)90021-2
Serru V Dessen P Boucheix C Rubinstein E Sequence and expression of seven new tetraspans Biochim Biophys Acta 2000 1478 159 163 10719184
Reichel C Brugger R Bang H Geisslinger G Brune K Molecular cloning and expression of a 2-arylpropionyl-coenzyme A epimerase; a key enzyme in the inversion metabolism of ibuprofen Mol Pharmacol 1997 51 576 582 9106621
Lee K Belinsky MG Bell DW Testa JR Krush GD Isolation of MOAT-B, a widely expressed multidrug resistance-associated protein/canalicular multispecific organic anion transporter-related transporter Cancer Res 1998 58 2741 2747 9661885
Nakajima-Jijima S Hamada H Reddy P Kakunaga T Molecular structure of human cytoplasmic beta-actin gene: interspecies homology of sequences in the introns Proc Natl Acad Sci USA 1985 82 6133 6137 2994062
Mubiru JN Shen-Ong GL Valente AJ Troyer DA Alternative spliced variants of alpha-methylacyl-CoA racemase gene and their expression in prostate cancer Gene 2004 327 89 98 14960364 10.1016/j.gene.2003.11.009
|
16143040
|
PMC1236915
|
CC BY
|
2021-01-04 16:03:06
|
no
|
BMC Cancer. 2005 Sep 5; 5:111
|
utf-8
|
BMC Cancer
| 2,005 |
10.1186/1471-2407-5-111
|
oa_comm
|
==== Front
BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1131615688810.1186/1471-2407-5-113Research ArticleTranslation elongation factor eEF1A2 is a potential oncoprotein that is overexpressed in two-thirds of breast tumours Tomlinson Victoria AL [email protected] Helen J [email protected] Naomi R [email protected] Juliette [email protected] Alexey [email protected] William R [email protected] J Michael [email protected] Catherine M [email protected] Medical Genetics, School of Molecular and Clinical Medicine, University of Edinburgh, Molecular Medicine Centre, Western General Hospital, Edinburgh EH4 2XU, UK2 Breast Unit Research Group, Western General Hospital, Edinburgh EH4 2XU, UK2005 12 9 2005 5 113 113 16 5 2005 12 9 2005 Copyright © 2005 Tomlinson 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 tissue-specific translation elongation factor eEF1A2 was recently shown to be a potential oncogene that is overexpressed in ovarian cancer. Although there is no direct evidence for an involvement of eEF1A2 in breast cancer, the genomic region to which EEF1A2 maps, 20q13, is frequently amplified in breast tumours. We therefore sought to establish whether eEF1A2 expression might be upregulated in breast cancer.
Methods
eEF1A2 is highly similar (98%) to the near-ubiquitously expressed eEF1A1 (formerly known as EF1-α) making analysis with commercial antibodies difficult. We have developed specific anti-eEF1A2 antibodies and used them in immunohistochemical analyses of tumour samples. We report the novel finding that although eEF1A2 is barely detectable in normal breast it is moderately to strongly expressed in two-thirds of breast tumours. This overexpression is strongly associated with estrogen receptor positivity.
Conclusion
eEF1A2 should be considered as a putative oncogene in breast cancer that may be a useful diagnostic marker and therapeutic target for a high proportion of breast tumours. The oncogenicity of eEF1A2 may be related to its role in protein synthesis or to its potential non-canonical functions in cytoskeletal remodelling or apoptosis.
==== Body
Background
Breast cancer is the most common cancer in females worldwide; there are an estimated 1 million new cases per year [1]. The identification of changes in gene expression in breast tumours relative to normal surrounding tissue is clearly of great importance in terms of prognostic indicators and therapeutic targets.
The translation elongation factor eEF1A2 was first identified as a tissue-specific variant of eEF1A1 (formerly known as EF-1α) in the early 1990s [2,3]. The two forms of eEF1A are encoded by separate loci, but the resulting proteins are 92% identical and 98% similar. Whereas eEF1A1 is widely expressed, eEF1A2 is normally expressed only in neurons and muscle [3-5]. The first specific evidence implicating eEF1A2 in tumourigenesis came in 2002 when Anand et al [6] showed that eEF1A2 was expressed in 30% of ovarian tumours, but not in normal ovary. The genomic region to which eEF1A2 maps, 20q13, had been known for many years to be amplified in a high proportion of ovarian and breast tumours [7]; [8], but the EEF1A2 gene maps closer to the telomere than the region previously implicated. Anand et al showed that 14/53 tumours had amplifications of the region surrounding EEF1A2 [6]. In the same paper, forced expression of eEF1A2 in cells was demonstrated to confer tumourigenic properties on NIH3T3 cells, and to give rise to tumours in xenografted nude mice. Although 20q13 amplification is commonly observed in breast cancer, there has as yet been no evidence for overexpression of eEF1A2 in breast tumours. However, eEF1A1, the widely-expressed isoform, was recently shown to be upregulated in the infiltrating edge of invasive breast tumours compared with the tumour core by microarray analysis of laser microdissected material, confirmed by immunohistochemistry [9]. In this analysis the antibody used was one that detects both eEF1A1 and eEF1A2 with equal intensity, so it is conceivable that eEF1A2 contributes to this pattern of expression.
We have generated antibodies (Newbery et al, in preparation) that allow us to distinguish between the highly related isoforms eEF1A1 (which is expressed in normal breast) and eEF1A2 (which is thought to be expressed only in muscle and neurons). Using these isoform-specific antibodies, we show that eEF1A2 expression is barely detectable in normal human breast tissue, but that the gene is moderately to strongly expressed in 63 % of breast tumours examined. Furthermore, there is a strong correlation between eEF1A2 overexpression and estrogen receptor (ER) positivity.
Methods
Quantitative Real-time Reverse Transcription-PCR (RT-PCR)
Breast cancer samples were obtained in the Edinburgh Breast Unit (Western General Hospital, Edinburgh) with patients' informed consent and ethical committee approval. Biopsies were snap frozen and stored in liquid nitrogen until RNA extraction. Before RNA extraction the frozen tissue was defrosted and stabilized in RNA-later-ICE reagent (Ambion). Total RNA was extracted with RNeasy-mini columns (Qiagen).
Amount and purity of RNA were evaluated by spectrophotometry. RNA integrity was confirmed by agarose gel electrophoresis. Total RNA was isolated from tumour and normal tissue using Qiagen RNeasy kits (Qiagen). RNA was treated with DNase using DNAfree kit (Ambion, Cambridgeshire) and 1 μg was used for RT-PCR using Retroscript kit (Ambion, Cambridgeshire, UK). TaqMan Assay-on-Demand gene expression pre-designed primer and probe sets from Applied Biosystems, Cheshire, UK were used for EEF1A2 (Assay # Hs 00157325 ml) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; control; Hs 99999905 ml). In a 10 μl reaction volume per well of a 394-well plate, 0.5 μl of primers, 5 μl of TaqMan PCR Master Mix, no AmpErase UNG 10×, and 4.5 μl of diluted cDNA were added (Applied Biosystems, Cheshire, UK). Real-time RT-PCR and the quantification of RT-PCR products were performed and the products analyzed using an ABI Prism 7900HT Sequence Detection System, and the appropriate software (SDS3.1) according to the manufacturer's instructions (Applied Biosystems, Cheshire, UK).
Western blots
Protein lysates from cell lines were prepared using previously published protocols [10]. Western blot analyses were carried out using standard protocols. The blots were incubated with primary anti-eEF1A2 rabbit antibody and primary anti-eEF1A1 sheep antibody diluted 1:200 in blocking solution, as well as primary anti-glyceraldehyde-3-phosphate dehydrogenase polyclonal mouse antibody (Chemicon International, Hampshire, UK) diluted 1:10000. Blots were then incubated in the appropriate horse radish peroxidase conjugated secondary antibody (Dako Cytomation, Cambridgeshire, UK) at 1:500. Detection was performed using enhanced chemiluminescence detection kit (Amersham Biosciences, Buckinghamshire, UK).
Immunohistochemistry
Specimens of normal and cancerous tumours were obtained with informed consent and local ethical committee approval from patients undergoing surgical treatment at the Royal Infirmary of Edinburgh and Western General Hospital, Edinburgh. A breast tumour histoarray (CB2) produced by SuperBioChips (AMS Biotechnology, Oxfordshire, UK) was also used. Formalin fixed, paraffin embedded sections of human normal tissue and tumour tissue were deparaffinized with xylene, rehydrated, treated with picric acid and microwaved in citric acid pH6. Slides were blocked in a 1:5 dilution of sheep serum for 30 minutes at room temperature. Primary anti-eEF1A2 rabbit antibodies were used at a concentration of 1:10 diluted in PBS, for 40 minutes at room temperature and secondary goat anti-rabbit IgG biotin conjugated antibody (Dako Cytomation, Cambridgeshire, UK) was used at 1:200 at room temperature for 30 minutes. Slides were incubated with StreptABC complex/HRP (Dako Cytomation, Cambridgeshire, UK) at room temperature for 30 minutes and in diaminobenzidene (Sigma Fast DAB, Sigma, Dorset, UK) for 2 minutes at room temperature. Finally slides were counterstained in haematoxylin, dehydrated and mounted in pertex.
Immunohistological scoring methods
The breast tumour sections and normal breast sections (CB2, SuperBioChips, AMS Biotechnology, Oxfordshire, UK) were scored as weak, moderate and strong staining for eEF1A2. Weak staining was considered as background since this level of staining is seen in normal tissue. Stromal tissue was negative in all cases. Blind scoring was carried out by two independent researchers. Two slides were analysed, representing different levels within tumours, and each of these was stained with a different antibody to eEF1A2. Almost perfect correlation was seen between the two slides.
Statistical methods
Fisher's exact test was used to test for associations between negative and weak eEF1A2 expressing tumours or moderately and strongly overexpressing tumours with ER positivity. For breast tumour Quantitative Real-time RT-PCR data, a two-sample t-test allowing for difference in variance between the two samples was used to test the difference between the mean standardised quantity of RNA for the ER-positive and ER-negative groups. P values that were less than or equal to 0.05 were considered significant.
Results
Expression analysis in breast tumours
Since 20q13.3 amplification is commonly seen in breast tumours as well as ovarian tumours, and because analysis of the SAGE database at NCBI indicated that eEF1A2 was more highly represented in breast tumours than in normal breast tissue (unpublished observations) we decided to examine eEF1A2 expression in breast tumours at both the RNA and protein level. Initially, we used Western blotting with an anti-eEF1A2 antibody to examine expression in a number of commonly-used cell lines. The anti-eEF1A2 antibody was raised against a peptide that differs significantly between eEF1A1 and eEF1A2 (Newbery et al, in preparation); specificity was confirmed by lack of signal from tissues taken from wasted mice which have a null mutation of eEF1A2 [5]. The majority of transformed cell lines showed high levels of expression of eEF1A2 (Figure 1); in addition, it has previously been shown that NIH3T3 cells do not express eEF1A2 except when they become confluent [2,6]. We therefore chose not to place any emphasis on the analysis of breast cancer cell lines as opposed to primary tumour samples since eEF1A2 expression seems to be a common property of transformed cells, rather than being specific for a tumour type. Instead, we carried out real-time quantitative RT-PCR of RNA samples from breast tumours. The results obtained are shown in Figure 2A. It can be seen that whereas extremely low levels of expression are detected in RNA samples from normal breast and from a benign breast tumour, most malignant tumour samples showed moderate to high (up to 30-fold higher than normal breast) expression levels. On average, the estrogen receptor (ER)-negative tumours showed only 1.2 times higher expression than the normal sample whereas ER-positive tumours had 8.4 times higher expression (Figure 2B). The difference in eEF1A2 expression levels between the estrogen receptor negative tumours and estrogen receptor positive tumours is 7.2 units (P = 0.0087, t-test, 95% confidence interval 2.0 to 12.4 units).
No protein extracts were available from these tumours for Western blot analysis so we then examined expression of eEF1A2 by immunohistochemistry on a commercial tissue array of normal and tumour breast samples using the anti-peptide antibodies described above. The array contained sections from 46 cases of cancer and 7 normal breasts; the results obtained are shown in Figure 3. None of the normal breast sections showed any more than faint staining. No stromal staining was observed and tumour staining within a sample was near-uniform. Of the tumour samples, 5 showed strong expression of eEF1A2 (11%) and 22 showed moderate expression (48%). The remainder appeared to have no more staining than normal breast. None of the three lobular carcinomas on the slide showed any eEF1A2 overexpression. There was no significant correlation between eEF1A2 expression level and tumour grade or lymph node positivity (data not shown). The tumours had all been previously assessed for p53 status; there appeared to be an association between overexpression of eEF1A2 and wild-type p53, but this was not statistically significant. The tendency to association may be a reflection of the significant association between ER positivity and p53 negativity (p = 0.012) in these tumour samples. Only four out of the 22 ER-negative tumours showed staining beyond background levels and none had strong staining. We found a significant association between eEF1A2 overexpression (scored as moderate/strong) and ER positivity (P = 0.016, Fisher's exact test). We then went on to examine 16 tissue sections from ER positive breast tumours obtained from patients at the Western General Hospital. Of these, 13 showed moderate or strong staining with the anti-eEF1A2 antibody. Overall then, 40 out of 63 breast tumours (63%) examined by immunohistochemistry showed significant overexpression of eEF1A2.
Discussion
We have shown that the putative oncogene eEF1A2 is upregulated in a high proportion of breast tumours. This upregulation is considerably more significant in ER-positive tumours. There is little or no detectable expression of eEF1A2 in normal breast tissue. It is not yet known whether this overexpression results from amplification of the EEF1A2 gene in all cases; in the study of ovarian tumours by Anand et al [6] at least one tumour showed overexpression in the absence of gene amplification, suggesting that there are other mechanisms by which the gene can be upregulated. There is a strong association between ER-positivity and eEF1A2 overexpression which is worthy of further study.
There also appears to be a weak correlation between the absence of p53 mutations and eEF1A2 overexpression. It is possible that eEF1A2 is not upregulated in tumours with p53 mutations because wild-type p53 is required for expression of eEF1A2 in certain cell types; it has been shown that p53 can upregulate expression of eEF1A1 [11], and the p53 binding sites identified in the gene encoding eEF1A1 are shared with that encoding eEF1A2 (unpublished observations). On the other hand it is conceivable that upregulation of eEF1A2 expression rather than p53 mutation is an alternative route for tumours to evade apoptosis in certain cancers.
The basis for the oncogenicity of eEF1A2 is still unclear. We, like Anand et al, have shown that the levels of eEF1A1 in tumours which over-express eEF1A2 are unchanged (data not shown), suggesting that these tumours might have a greater capacity for protein synthesis. However, it has been known for many years that eEF1A is in excess over the other components of the translation elongation apparatus [12], so eEF1A is unlikely to be rate-limiting in protein synthesis. eEF1A1 has been shown to determine the susceptibility of a number of independent cell lines to chemical- and UV-induced transformation [13] and has been identified as an actin binding protein in rat breast tumour cells, where it was found to be more highly expressed in metastatic than non-metastatic cells [14]. It is not yet clear whether these properties are shared with eEF1A2, but the availability of specific antibodies that distinguish between the two isoforms should allow us to shed light on this. One hypothesis is that the non-canonical ("moonlighting") properties of eEF1A1 [15] and eEF1A2 differ so that, for example, the way eEF1A2 interacts with the cytoskeleton might differ from that of eEF1A1 and affect the properties of cells which are expressing high levels of both isoforms. It has been shown, for example, that forced overexpression of eEF1A affects the cytoskeleton in both S. pombe and S. cerevisiae [16,17]. Alternatively, it has been shown that eEF1A1 and eEF1A2 differ in terms of their response to apoptotic agents [18]; the finding that eEF1A2 is anti-apoptotic, at least in certain conditions, has obvious implications for the possible role of eEF1A2 in tumourigenesis. The observation that eEF1A2 expression is seen in the majority of cell lines, regardless of the tissue of origin, suggests that eEF1A2 expression may be triggered by the general process of transformation. This idea is strengthened by the fact that most of the few cell lines which do not express eEF1A2 tend to be untransformed, such as NIH3T3 cells (Figure 1).
The presence of increased levels of eEF1A2 in breast tumours may provide a useful new diagnostic marker. Further, eEF1A2 may prove to be a feasible target for therapeutic intervention. It has already been shown that growth-factor mediated eEF1A1 expression can be blocked with anti-EGF antibodies [19]; it would be of interest to examine the response of eEF1A2 to similar antibodies. Investigations into non-canonical functions of eEF1A molecules may shed new light on mechanisms of oncogenicity.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
VT carried out the molecular analysis, Western blots and immunohistochemistry. HN isolated and characterised the antibodies. NW carried out the statistical analysis. AL made the RNA. JJ provided some of the tissue sections. WM and JMD collected the patient material. CA conceived the study and wrote the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We would like to thank Jeremy Thomas for discussion and analysis. This work was funded by Cancer Research UK and the Wellcome Trust.
Figures and Tables
Figure 1 Western blot analysis using an anti-eEF1A2 antibody on a range of cell lines. The loading control is GAPDH.
Figure 2 (A) Real-time RT-PCR analysis of RNA from breast tumours. Each block on the × axis represents a different tumour. The amount of eEF1A2 message is shown normalised to GAPDH and expressed relative to the level of expression in the normal breast RNA samples (=1). ER-negative tumours are shown in white, ER-positive tumours are shown in black. The difference in mean expression between ER-positive and ER-negative samples is 7.2 units (p = 0.0087), 95% Confidence Interval 2.0 to 12.4 units. (B) Average standardised RNA levels in ER-negative and ER-positive breast tumours. This difference is significant (P = 0.0087, t-test).
Figure 3 Immunohistochemistry of eEF1A2 in the breast. The panel labelled N shows the antibody staining weakly in a normal breast section. Panels T1 to T5 show breast tumours staining strongly with the anti-eEF1A2 antibody. Magnification ×10.
==== Refs
Baselga J Norton L Focus on breast cancer Cancer Cell 2002 1 319 322 12086846 10.1016/S1535-6108(02)00066-1
Ann DK Moutsatsos IK Nakamura T Lin HH Mao PL Lee MJ Chin S Liem RK Wang E Isolation and characterization of the rat chromosomal gene for a polypeptide (pS1) antigenically related to statin J Biol Chem 1991 266 10429 10437 1709933
Knudsen SM Frydenberg J Clark BF Leffers H Tissue-dependent variation in the expression of elongation factor-1 alpha isoforms: isolation and characterisation of a cDNA encoding a novel variant of human elongation-factor 1 alpha Eur J Biochem 1993 215 549 554 8354261 10.1111/j.1432-1033.1993.tb18064.x
Lee S Francoeur AM Liu S Wang E Tissue-specific expression in mammalian brain, heart, and muscle of S1, a member of the elongation factor-1 alpha gene family J Biol Chem 1992 267 24064 24068 1385435
Chambers DM Peters J Abbott CM The lethal mutation of the mouse wasted (wst) is a deletion that abolishes expression of a tissue-specific isoform of translation elongation factor 1alpha, encoded by the Eef1a2 gene Proc Natl Acad Sci U S A 1998 95 4463 4468 9539760 10.1073/pnas.95.8.4463
Anand N Murthy S Amann G Wernick M Porter LA Cukier IH Collins C Gray JW Diebold J Demetrick DJ Lee JM Protein elongation factor EEF1A2 is a putative oncogene in ovarian cancer Nat Genet 2002 31 301 305 12053177
Kallioniemi A Kallioniemi OP Piper J Tanner M Stokke T Chen L Smith HS Pinkel D Gray JW Waldman FM Detection and mapping of amplified DNA sequences in breast cancer by comparative genomic hybridization Proc Natl Acad Sci U S A 1994 91 2156 2160 8134364
Iwabuchi H Sakamoto M Sakunaga H Ma YY Carcangiu ML Pinkel D Yang-Feng TL Gray JW Genetic analysis of benign, low-grade, and high-grade ovarian tumors Cancer Res 1995 55 6172 6180 8521410
Zhu G Reynolds L Crnogorac-Jurcevic T Gillett CE Dublin EA Marshall JF Barnes D D'Arrigo C Van Trappen PO Lemoine NR Hart IR Combination of microdissection and microarray analysis to identify gene expression changes between differentially located tumour cells in breast cancer Oncogene 2003 22 3742 3748 12802281 10.1038/sj.onc.1206428
Gilmour LM Macleod KG McCaig A Sewell JM Gullick WJ Smyth JF Langdon SP Neuregulin expression, function, and signaling in human ovarian cancer cells Clin Cancer Res 2002 8 3933 3942 12473609
Kato MV Sato H Nagayoshi M Ikawa Y Upregulation of the elongation factor-1alpha gene by p53 in association with death of an erythroleukemic cell line Blood 1997 90 1373 1378 9269753
Slobin LI The role of eucaryotic factor Tu in protein synthesis. The measurement of the elongation factor Tu content of rabbit reticulocytes and other mammalian cells by a sensitive radioimmunoassay Eur J Biochem 1980 110 555 563 7192214 10.1111/j.1432-1033.1980.tb04898.x
Tatsuka M Mitsui H Wada M Nagata A Nojima H Okayama H Elongation factor-1 alpha gene determines susceptibility to transformation Nature 1992 359 333 336 1383827 10.1038/359333a0
Edmonds BT Wyckoff J Yeung YG Wang Y Stanley ER Jones J Segall J Condeelis J Elongation factor-1 alpha is an overexpressed actin binding protein in metastatic rat mammary adenocarcinoma J Cell Sci 1996 109 ( Pt 11) 2705 2714 8937988
Condeelis J Elongation factor 1 alpha, translation and the cytoskeleton Trends Biochem Sci 1995 20 169 170 7610475 10.1016/S0968-0004(00)88998-7
Munshi R Kandl KA Carr-Schmid A Whitacre JL Adams AE Kinzy TG Overexpression of translation elongation factor 1A affects the organization and function of the actin cytoskeleton in yeast Genetics 2001 157 1425 1436 11290701
Suda M Fukui M Sogabe Y Sato K Morimatsu A Arai R Motegi F Miyakawa T Mabuchi I Hirata D Overproduction of elongation factor 1alpha, an essential translational component, causes aberrant cell morphology by affecting the control of growth polarity in fission yeast Genes Cells 1999 4 517 527 10526238 10.1046/j.1365-2443.1999.00279.x
Ruest LB Marcotte R Wang E Peptide elongation factor eEF1A-2/S1 expression in cultured differentiated myotubes and its protective effect against caspase-3-mediated apoptosis J Biol Chem 2002 277 5418 5425 11724805 10.1074/jbc.M110685200
Talukder AH Jorgensen HF Mandal M Mishra SK Vadlamudi RK Clark BF Mendelsohn J Kumar R Regulation of elongation factor-1alpha expression by growth factors and anti-receptor blocking antibodies J Biol Chem 2001 276 5636 5642 11104760 10.1074/jbc.M006824200
|
16156888
|
PMC1236916
|
CC BY
|
2021-01-04 16:03:07
|
no
|
BMC Cancer. 2005 Sep 12; 5:113
|
utf-8
|
BMC Cancer
| 2,005 |
10.1186/1471-2407-5-113
|
oa_comm
|
==== Front
BMC Cardiovasc DisordBMC Cardiovascular Disorders1471-2261BioMed Central London 1471-2261-5-261615014310.1186/1471-2261-5-26Research ArticleThe "lipid accumulation product" performs better than the body mass index for recognizing cardiovascular risk: a population-based comparison Kahn Henry S [email protected] National Center for Chronic Disease Prevention and Health Promotion, CDC, Mail-stop K-10, 4770 Buford Highway, Atlanta, Georgia 30341-3717 USA2005 8 9 2005 5 26 26 2 3 2005 8 9 2005 Copyright © 2005 Kahn; licensee BioMed Central Ltd.2005Kahn; 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
Body mass index (BMI, kg/m2) may not be the best marker for estimating the risk of obesity-related disease. Consistent with physiologic observations, an alternative index uses waist circumference (WC) and fasting triglycerides (TG) concentration to describe lipid overaccumulation.
Methods
The WC (estimated population minimum 65 cm for men and 58 cm for women) and TG concentration from the third National Health and Nutrition Examination Survey (N = 9,180, statistically weighted to represent 100.05 million US adults) were used to compute a "lipid accumulation product" [LAP = (WC-65) × TG for men and (WC-58) × TG for women] and to describe the population distribution of LAP. LAP and BMI were compared as categorical variables and as log-transformed continuous variables for their ability to identify adverse levels of 11 cardiovascular risk factors.
Results
Nearly half of the represented population was discordant for their quartile assignments to LAP and BMI. When 23.54 million with ordinal LAP quartile > BMI quartile were compared with 25.36 million with ordinal BMI quartile > LAP quartile (regression models adjusted for race-ethnicity and sex) the former had more adverse risk levels than the latter (p < 0.002) for seven lipid variables, uric acid concentration, heart rate, systolic and diastolic blood pressure. Further adjustment for age did not materially alter these comparisons except for blood pressures (p > 0.1). As continuous variables, LAP provided a consistently more adverse beta coefficient (slope) than BMI for nine cardiovascular risk variables (p < 0.01), but not for blood pressures (p > 0.2).
Conclusion
LAP (describing lipid overaccumulation) performed better than BMI (describing weight overaccumulation) for identifying US adults at cardiovascular risk. Compared to BMI, LAP might better predict the incidence of cardiovascular disease, but this hypothesis needs prospective testing.
==== Body
Background
Obesity is commonly understood to imply excess fat, but it is ordinarily classified according to excess weight. This semantic inconsistency may help to explain why the body mass index (BMI, kg/m2) – a popular marker for relative weight – performs only modestly as a predictor of medical risk [1]. Researchers have increasingly appreciated that adipose tissue has complex functions [2,3], and that these functions may vary according to anatomic region [4-8]. Some of the region-specific functions of adipose tissue are beneficial, such as storing and buffering daily fluxes of circulating lipid fuels [9]. Thus, one cannot assume that a high relative weight or global adiposity is always deleterious.
In the current era of increasing obesity, we should attempt to define and measure lipid accumulation specifically in those contexts where accumulation may represent a physiologic danger [10,11]. These contexts might be described as lipid overaccumulation [12]. At the same time we should avoid attributing culpability to components of adipose or lean tissue that are enlarged but might enhance physiologic processes or reduce the risk of disease.
This paper first describes a simple index for estimating lipid overaccumulation among adults. Next, to demonstrate the utility of the lipid overaccumulation concept, this paper tests the hypothesis that the described index is better correlated than BMI with a variety of cardiovascular risk factors. This hypothesis should not be surprising since the BMI can neither distinguish between fat and lean tissues nor identify the anatomic location or function of distinct fat depots.
The proposed index – designated the "lipid accumulation product" (LAP) – is based on a combination of two measurements that are safe and inexpensive to obtain. One is waist circumference (WC), a measure of truncal fat that includes the visceral (intra-abdominal) depot. The other is the fasting concentration of circulating triglycerides (TG), the esterified, long-chain fatty acids that circulate through blood contained stably inside lipoproteins. Both waist size and TG concentration tend to rise with age [13,14], suggesting that their values are subject to accumulation over time. Waist size and circulating TGs are each continuously associated with metabolic insulin resistance [15,16], daylong triglyceridemia [17], and cardiovascular risk [18-20].
Although previous papers have proposed that the combination of enlarged waist and elevated TGs might serve as a dichotomous risk marker [21,22], the simple index described here was developed to express a continuous risk function. If LAP is better correlated than BMI with cardiovascular risk factors, this finding would support the notion that the overaccumulation of lipid carries worse cardiovascular consequences than the less specific overaccumulation of weight.
Methods
Defining the lipid accumulation product (LAP)
Data were obtained from the third National Health and Nutrition Examination Survey (NHANES III), a probability sample of the US civilian, noninstitutionalized population that included an oversample of non-Hispanic blacks and Mexican Americans [23]. NHANES III, conducted over the period 1988–1994, was unique among large US surveys because it incorporated assays of serum apolipoproteins B and A1 (ApoB, ApoA1; restricted to survey years 1988–1991). The analytic population from the seven-year survey contained 4,447 male and 4,733 female participants who were aged 18+ years, not pregnant, had fasted 8–19 hours before their laboratory examination, and had data available on basic anthropometry and fasting serum TGs (excluding three persons with a TG concentration >15 mmol/L). Participants were asked to complete a household interview and a standardized examination, including measurement of the standing waist circumference (in the horizontal plane at the level just above the right iliac crest, at minimal respiration) [24,25]. Serum TGs were measured enzymatically after hydrolysis to glycerol (Hitachi 704 Analyzer; Boehringer Mannheim, Indianapolis, Indiana); the coefficient of variation was 3–5 percent over the study and across the clinical range. Additional details of all laboratory procedures are available elsewhere [26]. Calculations of low-density lipoprotein (LDL) cholesterol concentration were limited to participants with TG concentrations below 4.5 mmol/L (a requirement of the Friedewald equation [27]) who had fasted at least nine hours.
Sampling weights from NHANES III were used with the software programs SAS, SAS/Graph (Release 8.2, SAS Institute, Cary, NC), and SUDAAN (Release 8.0, Research Triangle Institute, Research Triangle, NC), to estimate the sizes of the represented adult populations, to describe the distributions in the population of risk factors associated with LAP and BMI, and to perform analyses using multivariable linear regression. The analyses thus incorporated sampling weights that accounted for unequal selection probabilities (clustered design, planned oversampling, and differential nonresponse) [28]. Based on the sampling weights assigned, the analytic cohort represented an estimated total of 100,048,439 US adults aged 18+ years, 50.5 percent (SE 0.7) of them women, with a distribution of race-ethnicity that was 76.0 (1.5) percent non-Hispanic white, 10.5 (0.6) percent non-Hispanic black, 5.3 (0.5) percent Mexican American, and 8.1 (1.1) percent other.
Sex-specific bubble plots of population density by the values for WC (to the nearest cm) and TG concentration (to the nearest 0.1 mmol/L) were prepared to represent US adults in three age ranges (figure 1). The area of each bubble on these plots is proportional to the estimated number of men or women represented by those intersections. A sex-specific hypothetical minimum value for WC (that is, the waist size that theoretically contained only abdominal muscle, viscera, and vertebral bone) was estimated by calculating the mean minus two standard deviations of the log-transformed WC value among the estimated 15.00 million persons aged 18–24 years. These estimated minimum WC values (65 cm for men and 58 cm for women) were very similar to minimum values reported in a survey of 18 year-old Canadians in 1981 [29].
Figure 1 Age-specific distributions by waist circumference and fasting triglyceride concentrations, from NHANES III. For consecutive age groups the population density is increasingly displaced away from the hypothetical origin points of zero lipid accumulation (solid triangles). For clarity of presentation, the bubble plots omit the extreme outlying values for WC and TG.
The minimum WC values were used to define sex-specific origin points (near the left-lower corner of each panel in figure 1) that represented a hypothetical state in which TG concentrations were arbitrarily set to zero and the waist size (greater for men than women) comprised primarily lean truncal tissue. A comparison of sex-specific bubble plots by age group confirmed that increasing age was accompanied by a shift of the population density upward (increasing waist size) and to the right (increasing circulating TG). Consistent with this empirical observation, the LAP was defined to describe the extent to which an individual had travelled the route – in theory – of both increasing waist and increasing TG:
LAP for men = (WC [cm] - 65) × (TG concentration [mmol/L])
LAP for women = (WC [cm] - 58) × (TG concentration [mmol/L])
In order to avoid having nonpositive values for LAP, any waist values for men that were 65 cm or less (five men in the NHANES III sample, all aged 18–22 years) were revised upward to 66.0 cm. No women in the entire NHANES III sample had a waist circumference less than 58.4 cm.
Comparing LAP with BMI
Assessment of discordant subpopulations
Starting with the estimated full population, two subpopulations were identified for which the quartile classifications for LAP and BMI were discordant (table 1). The subpopulation whose ordinal LAP quartile was greater than their ordinal BMI quartile (i.e., those located below the bolded diagonal cells in table 1) was compared with the subpopulation whose ordinal LAP quartile was less than their ordinal BMI quartile (i.e., those located above the bolded diagonal cells). An estimated 48.91 million individuals (47.7 percent of the men, 50.1 percent of the women) were located in discordant quartiles. To calculate the effect of these discordant classifications on 11 cardiovascular risk variables, the two discordant subpopulations were compared in linear regression models using adjustments for sex and race-ethnicity and then again with further adjustment for age (terms for age and age2). All comparisons are reported with two-sided p values.
Table 1 Distribution of US adults by population quartiles of lipid accumulation product and body mass index. Table shows number of survey participants and corresponding population estimates (millions, in parentheses). Participants identified in the bold-print cells are concordant for their quartile assignment both to LAP and BMI.
Quartiles of lipid accumulation product (LAP)
4 3 2 1 Total
Quartiles of body mass index (BMI) 4 1543 (15.45) 809 (7.30) 248 (2.26) 21 (0.17) 2621 (25.19)
3 686 (6.93) 983 (9.66) 724 (7.28) 164 (1.42) 2557 (25.30)
2 191 (2.23) 531 (6.08) 742 (9.53) 521 (6.93) 1985 (24.77)
1 34 (0.39) 176 (1.95) 494 (5.97) 1313 (16.49) 2017 (24.79)
Total 2454 (25.00) 2499 (24.99) 2208 (25.03) 2019 (25.02) 9180 (100.05)
Comparison of continuous linear regression models in the full population
Linear regression models were prepared from the entire population (discordant and concordant) that used either log-transformed LAP (ln LAP) or log-transformed BMI (ln BMI) as continuous independent variables. The dependent (outcome) variables were the same set of 11 cardiovascular risk factors. Models were also prepared separately for two age groups, those under age 50 years and those aged 50+ years, with adjustments for race-ethnicity and for sex when the sexes were combined. For each outcome risk variable, ln LAP and ln BMI were evaluated by comparing the proportion of the total variation that each index could explain, that is, R2 for the entire model minus R2 for a base model that excluded ln LAP and ln BMI. For these continuous analyses, the beta coefficients (slopes) were standardized to reflect the increment in each outcome variable associated with an increment of one standard deviation from the mean (calculated for each sex and age group [18–49 years or 50+ years]) of either ln LAP or ln BMI.
Results
Distributions of LAP and BMI
The population-based distributions of LAP and BMI were skewed to the right (table 2), but with logarithmic transformation both indices approached a normal distribution within each age group. Mean and median values rose consistently with age for both LAP and BMI. However, for LAP (more than for BMI) the values for men rose more rapidly before age 50 years. Above that age the sex differences were attenuated. Within the middle age range (25–49 years) the upper quartile of men's LAP was greater than the women's upper quartile, but for the upper quartile of BMI the sex contrast went in the opposite direction.
Table 2 Population estimates of lipid accumulation product and body mass index by sex and age group. Estimates derived for US adults from NHANES III, 1988–1994.
Lipid accumulation product (LAP) cm·mmol/L Body mass index (BMI) kg/m2
P e r c e n t i l e P e r c e n t i l e
Sex & age Survey sample (N) Geometric mean 25th 50th 75th Geometric mean 25th 50th 75th
Men
18–24 years 685 16.2 9.2 15.5 27.6 23.6 21.1 23.1 25.7
25–49 years 1982 35.0 20.1 35.5 63.2 26.4 23.7 25.9 29.0
50+ years 1780 52.4 33.3 53.4 85.6 26.9 24.3 26.8 29.9
Women
18–24 years 715 16.6 9.4 16.0 27.6 23.1 20.0 22.4 25.6
25–49 years 2216 25.7 13.6 24.6 47.7 25.4 21.5 24.6 29.5
50+ years 1802 50.2 29.6 51.7 84.5 26.9 23.1 26.4 30.7
All
18–24 years 1400 16.4 9.2 15.7 27.6 23.4 20.6 22.8 25.6
25–49 years 4198 30.1 16.4 29.7 57.3 25.9 22.5 25.3 29.1
50+ years 3582 51.2 31.1 52.6 85.3 26.9 23.8 26.6 30.3
Applied to the entire adult age range, the male and female quartile cutpoints for LAP were similar, although the men's values were slightly higher than the women's (figure 2). For BMI, the quartile cutpoints were higher for men than women at the 25th and 50th percentiles (23.3 and 25.7 vs. 21.7 and 24.8 kg/m2), but at the 75th percentile the BMI cutpoint was lower for men than women (28.9 vs. 29.6 kg/m2).
Figure 2 Lines of equivalent percentile value for the lipid accumulation product (LAP) among US adults. Population estimates from NHANES III (USA, 1988–1994) are shown separately for men (top panel) and women (bottom panel). The presented iso-LAP values (95th through 25th percentiles) are 144.7, 112.0, 66.1, 37.4, and 19.1 cm·mmol/L for men and 135.6, 103.5, 60.4, 30.3, and 15.6 cm·mmol/L for women.
The linear correlation between LAP and BMI for the adult population was modest (r = 0.58) and somewhat stronger when both indices were log-transformed (r = 0.71).
Comparisons of LAP with BMI regarding cardiovascular risk variables
Analyses restricted to discordant subpopulations
Compared to the subpopulation with ordinal BMI quartile > ordinal LAP quartile, the subpopulation with ordinal LAP quartile > ordinal BMI quartile was older [50.5 (SE 0.7) years vs. 38.5 (0.6) years], had more non-Hispanic whites [82.2 (1.8) vs. 69.4 (2.0) percent], and fewer non-Hispanic blacks [5.2 (0.5) vs. 17.5 (1.3) percent] and Mexican Americans [4.4 (0.5) vs. 5.9 (0.6) percent]. In regression models adjusted for sex and race-ethnicity (table 3), they had more adverse levels for all 11 of the evaluated cardiovascular risk factors (p < 0.002). These included the two factors for which mean values are inversely associated with risk – concentration of high-density-lipoprotein (HDL) cholesterol and ratio of low-density-lipoprotein (LDL) cholesterol to apolipoprotein B (i.e., smaller size of LDL particles [30]). With additional adjustment for age, the relatively adverse status of higher LAP was no longer seen for systolic and diastolic blood pressure (p > 0.1), but it remained (p < 0.0005) for all seven lipid variables, uric acid concentration, and heart rate (table 3, right columns). The exclusion of persons who reported that they were taking medicine prescribed for high blood pressure (prevalence 11.3 percent) or to lower their cholesterol (prevalence 2.8 percent) did not alter any of these observed relationships (data not shown).
Table 3 Mean levels of cardiovascular risk variables among the subpopulations discordant for quartiles of LAP and BMI. Estimates derived for US adults from NHANES III, 1988–1994.
Mean (SE) adjusted for sex and race-ethnicity Mean (SE) adjusted for sex, race-ethnicity, and age
Dependent variable, units Discordant survey sample (N) LAP quartile > BMI quartile BMI quartile > LAP quartile P value LAP quartile > BMI quartile BMI quartile > LAP quartile P value
Total cholesterol, mmol/L 4598 5.62 (0.04) 4.89 (0.04) <0.0001 5.50 (0.04) 5.00 (0.04) <0.0001
HDL cholesterol, mmol/L 4578 1.24 (0.01) 1.37 (0.01) <0.0001 1.22 (0.01) 1.39 (0.01) <0.0001
LDL cholesterol, mmol/L 3333 3.50 (0.04) 3.11 (0.05) <0.0001 3.40 (0.04) 3.20 (0.05) 0.0003
Total cholesterol / HDL cholest. 4577 4.93 (0.06) 3.78 (0.05) <0.0001 4.88 (0.06) 3.82 (0.05) <0.0001
Apolipoprotein B, g/L 2245* 1.13 (0.01) 0.94 (0.01) <0.0001 1.10 (0.01) 0.97 (0.01) <0.0001
ApoB/ApoA1 2230* 0.809 (0.012) 0.681 (0.010) <0.0001 0.799 (0.013) 0.690 (0.010) <0.0001
LDL cholesterol /ApoB, mmol/g 1614* 3.06 (0.04) 3.26 (0.03) 0.0007 3.03 (0.04) 3.29 (0.03) <0.0001
Uric acid, mmol/L 4533 327 (2) 311 (3) <0.0001 325 (2) 313 (3) 0.0004
Systolic blood pressure, mmHg 4595 124.9 (0.7) 118.3 (0.4) <0.0001 121.3 (0.6) 121.6 (0.4) 0.62
Diastolic blood pressure, mmHg 4595 74.5 (0.3) 73.0 (0.3) 0.0018 74.1 (0.3) 73.4 (0.3) 0.14
Heart rate, bpm 4494 75.0 (0.5) 72.4 (0.6) 0.0005 75.1 (0.5) 72.3 (0.6) 0.0001
* Lipoprotein data obtained only during phase 1 of NHANES III.
Analyses of continuous linear regression models in the full population
Ln LAP consistently explained a greater portion of the variation of the outcome variables than did ln BMI for all seven lipid outcome variables, uric acid concentration, and heart rate (figure 3). For most of these variables the proportion of total variance explained (R2) by ln LAP was about twice that of ln BMI. For the remaining two variables, systolic and diastolic blood pressure, the contrasts in portion of explained variation were small but consistent between the sexes. In the population under age 50 years, ln BMI was the better predictor of systolic blood pressure, but ln LAP was better for diastolic pressure. In the population aged 50+ years, ln LAP was the stronger predictor of systolic blood pressure.
Figure 3 Proportion of total population variation (R2) in risk variables explained by Ln LAP and Ln BMI. Histograms were estimated from NHANES III data (USA,1988–1994) showing R2 values for sex-specific, age-specific, regression models of cardiovascular risk variables after adjustment for race-ethnicity.
A similar pattern was seen in a comparison of the standardized beta coefficients applied to the entire adult age range (data not shown). The beta coefficient (slope) of ln LAP was consistently more adverse (p < 0.01) than that of ln BMI for nine of the cardiovascular risk variables, but not for systolic and diastolic blood pressures (p > 0.2). When these comparative models were stratified by the two age groups the relatively greater slope of ln LAP was preserved among the nine variables for both age groups (table 4). Again, the exclusion of self-reported medication users did not alter these relationships (data not shown).
Table 4 Estimated increments in each risk variable per 1 standard deviation of ln LAP or ln BMI. Population estimates derived for US adults from NHANES III (1988–1994) with adjustments for sex and race-ethnicity.
Dependent variable Age (years) range Mean outcome value, units Observations (N) used in analyses Increment (SE) per 1 SD in ln LAP Increment (SE) per 1 SD in ln BMI
Total cholesterol 18–49 4.96 mmol/L 5597 0.413 (0.023) 0.225 (0.018)***
50+ 5.75 mmol/L 3582 0.320 (0.024) 0.089 (0.022)***
HDL cholesterol 18–49 1.29 mmol/L 5571 -0.152 (0.008) -0.111 (0.007)**
50+ 1.32 mmol/L 3570 -0.187 (0.009) -0.109 (0.010)***
LDL cholesterol 18–49 3.09 mmol/L 3858 0.326 (0.023) 0.220 (0.021)**
50+ 3.62 mmol/L 2754 0.146 (0.031) 0.058 (0.026)
Total cholesterol/HDL cholest. 18–49 4.16 mmol/L 5570 0.905 (0.040) 0.584 (0.033)***
50+ 4.74 mmol/L 3570 0.930 (0.042) 0.383 (0.042)***
Apolipo-protein B 18–49 0.97 g/L 2668† 0.123 (0.006) 0.074 (0.007)***
50+ 1.16 g/L 1807† 0.115 (0.010) 0.056 (0.007)***
ApoB/ApoA1 18–49 0.717 2642† 0.115 (0.006) 0.074 (0.006)***
50+ 0.810 1801† 0.109 (0.009) 0.069 (0.006)**
LDL cholesterol/ApoB 18–49 3.15 mmol/g 1813† -0.090 (0.017) -0.013 (0.012)**
50+ 3.19 mmol/g 1393† -0.160 (0.018) -0.066 (0.022)**
Uric acid 18–49 315 μmol/L 5524 27 (1) 25 (1)
50+ 334 μmol/L 3506 30 (2) 23 (2)*
Systolic blood pressure 18–49 115.1 mm Hg 5594 3.8 (0.2) 3.9 (0.2)
50+ 134.1 mm Hg 3580 3.1 (0.3) 2.0 (0.4)
Diastolic blood pressure 18–49 72.9 mm Hg 5592 3.7 (0.2) 3.3 (0.2)
50+ 75.5 mm Hg 3580 2.0 (0.2) 2.0 (0.2)
Heart rate 18–49 73.4 bpm 5482 2.4 (0.3) 1.8 (0.2)
50+ 74.1 bpm 3493 2.2 (0.3) 1.2 (0.3)
* p < 0.01, ** p < 0.001, *** p < 0.0001, indicating significantly different beta coefficients for ln LAP and ln BMI.
† Lipoprotein data obtained only during phase 1 of NHANES III.
When the separate component measures that contribute to LAP (i.e., WC and TG concentration) were log-transformed and entered individually into predictive models, their standardized beta coefficients generally showed lesser or equivalent (p > 0.05) slopes compared to the slopes for LAP. The only exception was for the estimation of the ratio LDL cholesterol/Apo B among adults 18–49 years old. In this group the slope for ln TG alone [-0.160 (0.019)] was more steeply negative (p = 0.02) than the slope for ln LAP [-0.090 (0.017)].
Discussion
The index described in this paper – the lipid accumulation product (LAP) – was developed in an effort to reflect the combined anatomic and physiologic changes associated with lipid overaccumulation in adults. Compared with BMI, LAP exhibited better correlations with lipid risk variables, uric acid concentration, and heart rate, but its correlation with blood pressure was roughly equivalent.
It is reasonable to speculate that the two LAP components – that is, enlarged abdominal fat depots and increased TG concentration – are each an indication that available lipid fuels have exceeded the individual's capacity to buffer and safely store this major form of acquired energy. Prior to 50 years old, the LAP appears to rise more slowly with age for women compared to men (Table 2). The women's relative delay of lipid overaccumulation is consistent with their greater amount of lower-body adipose tissue that confers increased buffering and storage capacity.
Whether an individual's excess lipid fuel appears eventually as an enlarged abdomen or as elevated circulating TG could be dictated in part by genes or by features of the individual's environmental circumstances. A special case, by way of an extreme example, might be the rare individual who is genetically disposed to extremely high TG concentrations (chylomicronemia). For the purpose of risk assessment in the general adult population, however, the alternative manifestations of lipid overaccumulation could be similarly informative. Regardless if the overaccumulation is marked by waist size, by TG concentration, or by both, the calculated value of LAP will be increased. In parallel with the LAP increments, excess lipid material will increasingly be deposited in nonadipose, "ectopic" tissues (e.g., liver, skeletal muscle, heart, blood vessels, kidneys, and pancreas) where it may adversely modify cellular metabolism, accelerate apoptosis (cell death), and interfere with cardiovascular control [10,11,31].
Ectopic lipid deposition is difficult to quantify directly, but an increased LAP value may indicate that various tissues or organs have become more vulnerable to injury from lipid overaccumulation. Lipoprotein particles with small diameters are more associated with disease risk than those with large diameters [32,33], and lipoprotein particle diameter is inversely associated with abdominal size [34]. Thus, LAP may effectively express disease risk through the hyperbolic relationship between LAP and its two component variables (Figure 2). The presence of an enlarged waist (implying a small lipoprotein particle size) allows the LAP value to increase rapidly with each unit increase in TG concentration. By contrast, the presence of a small waist (implying a large lipoprotein particle size) allows the LAP value to increase slowly with each unit increase in TG concentration. Although the NHANES III data set contains no direct measurement of particle sizes, our indirect estimate of LDL particle size (LDL cholesterol/ApoB) confirms that a small LDL particle size is better correlated with LAP than with BMI (Figure 3 and Table 4).
In contrast to an elevation in LAP value, an elevated BMI value (i.e., relative weight) is less specific in its anatomic or physiologic implications. Increased weight might represent enhancement of lean tissues, enlargement of the protective, subcutaneous adipose depots in the lower extremities [5,35-37], or systemic overload of fluid – changes that could be either salutary or simply secondary consequences of other disease processes. The commonly observed association of fluid overload with hypertension may explain the instances in which this study found blood pressure to be marginally better correlated with BMI than with LAP (Table 4 and Figure 3).
In order for LAP to gain a useful role in clinical medicine or epidemiology, at least three major questions remain to be addressed:
1. Is LAP strongly predictive of major disease outcomes? This preliminary analysis examined only intermediary outcomes (risk factors), and its data are entirely cross-sectional. However, a 20-year followup study of Swedish women reported that abdominal adiposity and elevated TG concentration were associated with increased risks of death from myocardial infarction and all causes, but that elevated BMI and cholesterol concentration were much weaker predictors [38]. In another prospective study from Scandinavia, post-menopausal women followed for 8.5 years demonstrated that the baseline combination of enlarged waist with elevated triglycerides – a dichotomous marker – was a very strong predictor of all-cause and cardiovascular mortality as well as the annual progression rate of aortic calcification [39]. More data collected prospectively in a variety of populations, including men, would help confirm that LAP – a continuous marker – has advantages over other simple indices for predicting the incidence of major diseases and mortality.
2. Is it useful to monitor LAP values as an indicator of intervention effectiveness? A recent report of intentional weight loss among overweight and obese Japanese women found that their exercise regimen and low-calorie diet were rewarded with a 37 percent reduction in TG concentration and a 27 percent reduction in truncal fat but only a 2 percent reduction in leg fat and a 12 percent reduction in BMI [40]. The authors also reported a direct correlation between changes in truncal fat and changes in fasting TG concentration or the number of heart disease risk factors, but that the changes in leg fat were inversely correlated with the changes in TG concentration or the number of heart disease risk factors. In another study, Italian men and women with diabetes were followed for two years after randomization to a physical activity counselling intervention [41]. Across six levels of aerobic energy expenditure, those who exercised more experienced significant reductions (p < 0.001) in waist circumference and circulating TG, but no reductions (p > 0.25) in either weight or BMI. Thus, the participants in both of these studies would have achieved a substantial reduction in LAP in association with improved cardiovascular risk factors, but their reduction in BMI was modest and less clearly associated with cardiovascular benefit. These observations from Asia and Europe demonstrate a potential advantage to using LAP as an intermediary variable by which to assess interventions against obesity-related risk.
3. Is LAP a practical index for adoption by clinicians or epidemiologists? Standardized waist measurements are highly reproducible [42], and they are arguably simpler and less expensive to obtain than well standardized weights and heights. The requirement of a venipuncture, however, could be a major obstacle for many potential participants or patients. The need for the fasting state could also represent a major inconvenience, although many persons accept the fasting condition when necessary for assessment of their glucose or lipid status. Of special importance in less developed economies, the laboratory cost of a single assay for TGs would be low compared with the costs of multiple assays for lipoproteins and an extensive biochemical panel.
Conclusion
The cross-sectional associations with LAP demonstrated in this paper should be seen primarily as a demonstration of how the concept of lipid overaccumulation may be expressed in an adult population. The utility of LAP for research or as a practical tool for use in the community will depend on the degree to which LAP can be demonstrated to enhance prediction of disease incidence. Prospective data sets that include baseline information on WC and fasting TG concentration would be well suited to evaluate LAP as a predictor of cardiovascular outcomes and mortality.
Abbreviations
ApoA1, apolipoprotein A1
ApoB, apolipoprotein B
BMI, body mass index
HDL, high-density lipoprotein
LAP, lipid accumulation product
LDL, low density lipoprotein
NHANES III, third National Health and Nutrition Examination Survey
Competing interests
The author(s) declares that he has no competing interests.
Authors' contributions
HK conceived of this study, performed the calculations, and drafted the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The author acknowledges the extraordinary efforts of the field staff, laboratory personnel, and statisticians who collected and processed the information in NHANES III.
==== Refs
McGee DL Body mass index and mortality: a meta-analysis based on person-level data from twenty-six observational studies Ann Epidemiol 2005 15 87 97 15652713 10.1016/j.annepidem.2004.05.012
Frayn KN Karpe F Fielding BA MacDonald IA Coppack SW Integrative physiology of human adipose tissue Int J Obes Relat Metab Disord 2003 27 875 888 12861227 10.1038/sj.ijo.0802326
Havel PJ Update on adipocyte hormones: Regulation of energy balance and carbohydrate/lipid metabolism Diabetes 2004 53 S143 S151 14749280
Lamarche B Abdominal obesity and its metabolic complications: implications for the risk of ischaemic heart disease Coron Artery Dis 1998 9 473 481 9847978
Van Pelt RE Evans EM Schechtman KB Ehsani AA Kohrt WM Contributions of total and regional fat mass to risk for cardiovascular disease in older women Am J Physiol Endocrinol Metab 2002 282 E1023 E1028 11934666
Tanko LB Bagger YZ Alexandersen P Larsen PJ Christiansen C Central and peripheral fat mass have contrasting effect on the progression of aortic calcification in postmenopausal women Eur Heart J 2003 24 1531 1537 12919778 10.1016/S0195-668X(03)00319-1
Ferreira I Snijder MB Twisk JWR van Mechelen W Kemper HCG Seidell JC Stehouwer CDA Central fat mass versus peripheral fat and lean mass: opposite (adverse versus favorable) associations with arterial stiffness? the Amsterdam Growth and Health Longitudinal Study J Clin Endocrinol Metab 2004 89 2632 2639 15181034 10.1210/jc.2003-031619
Karelis AD St Pierre DH Conus F Rabasa-Lhoret R Poehlman ET Metabolic and body composition factors in subgroups of obesity: what do we know? J Clin Endocrinol Metab 2004 89 2569 2575 15181025 10.1210/jc.2004-0165
Frayn KN Adipose tissue as a buffer for daily lipid flux Diabetologia 2002 45 1201 1210 12242452 10.1007/s00125-002-0873-y
Schaffer JE Lipotoxicity: when tissues overeat Curr Opin Lipidol 2003 14 281 287 12840659 10.1097/00041433-200306000-00008
Unger RH Weapons of lean body mass destruction: the role of ectopic lipids in the metabolic syndrome Endocrinology 2003 144 5159 5165 12960011 10.1210/en.2003-0870
Unger RH Lipid overload and overflow: metabolic trauma and the metabolic syndrome Trends Endocrinol Metab 2003 14 398 403 14580758 10.1016/j.tem.2003.09.008
Dobbelsteyn CJ Joffres MR MacLean DR Flowerdew G A comparative evaluation of waist circumference, waist-to-hip ratio and body mass index as indicators of cardiovascular risk factors. The Canadian Heart Health Surveys Int J Obes Relat Metab Disord 2001 25 652 61 11360147 10.1038/sj.ijo.0801582
MacLean DR Petrasovits A Connelly PW Joffres M O'Connor B Little JA Plasma lipids and lipoprotein reference values, and the prevalence of dyslipoproteinemia in Canadian adults. Canadian Heart Health Surveys Research Group Can J Cardiol 1999 15 434 444 10322253
Kriketos AD Furler SM Gan SK Poynten AM Chisholm DJ Campbell LV Multiple indexes of lipid availability are independently related to whole body insulin action in healthy humans J Clin Endocrinol Metab 2003 88 793 798 12574215 10.1210/jc.2002-020848
Carr DB Utzschneider KM Hull RL Kodama K Retzlaff BM Brunzell JD Shofer JB Fish BE Knopp RH Kahn SE Intra-abdominal fat is a major determinant of the National Cholesterol Education Program Adult Treatment Panel III criteria for the metabolic syndrome Diabetes 2004 53 2087 2094 15277390
van Wijk JP Halkes CJ Erkelens DW Castro CM Cabezas Fasting and daylong triglycerides in obesity with and without type 2 diabetes Metabolism 2003 52 1043 1049 12898471 10.1016/S0026-0495(03)00106-9
Jeppesen J Hein HO Suadicani P Gyntelberg F Triglyceride concentration and ischemic heart disease: an eight-year follow-up in the Copenhagen Male Study Circulation 1998 97 1029 1036 9531248
Austin MA Hokanson JE Edwards KL Hypertriglyceridemia as a cardiovascular risk factor Am J Cardiol 1998 81 7B 12B 9462597 10.1016/S0002-9149(98)00031-9
Janssen I Katzmarzyk PT Ross R Waist circumference and not body mass index explains obesity-related health risk Am J Clin Nutr 2004 79 379 384 14985210
Lemieux I Pascot A Couillard C Lamarche B Tchernof A Almeras N Bergeron J Gaudet D Tremblay G Prud'homme D Nadeau A Despres JP Hypertriglyceridemic waist: A marker of the atherogenic metabolic triad (hyperinsulinemia; hyperapolipoprotein B; small, dense LDL) in men? Circulation 2000 102 179 184 10889128
Kahn HS Valdez R Metabolic risks identified by the combination of enlarged waist and elevated triacylglycerols Am J Clin Nutr 2003 78 928 934 14594778
National Center for Health Statistics (Hyattsville MD) Plan and Operation of the Third National Health and Nutrition Examination Survey, 1988-1994 1994
Centers for Disease Control and Prevention National Health and Nutrition Examination Survey: Body Measurements (Anthropometry) 1988 1988 Atlanta, GA: National Center for Health Statistics
Chumlea NC Kuczmarski RJ Using a bony landmark to measure waist circumference (letter) J Am Diet Assoc 1995 95 12. 7798573 10.1016/S0002-8223(95)00003-8
Gunter EW Lewis BG Koncikowski SM Laboratory Procedures Used for the Third National Health and Nutrition Examination Survey (NHANES-III), 1988-1994 1996 Atlanta, GA: Centers for Disease Control and Prevention
Friedewald WT Levy RI Fredrickson DS Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge Clin Chem 1972 18 499 502 4337382
Mohadjer L Montaquila J Waksberg J Bell B James P Flores-Cervantes I Montes M National Health and Nutrition Examination Survey III, Weighting and Estimation Methodology 1996 Hyattsville, MD: National Center for Health Statistics
Katzmarzyk PT Waist circumference percentiles for Canadian youth 11-18 y of age Eur J Clin Nutr 2004 58 1011 1015 15220942 10.1038/sj.ejcn.1601924
Wagner AM Jorba O Rigla M Alonso E Ordonez-Llanos J Perez A LDL-cholesterol/apolipoprotein B ratio is a good predictor of LDL phenotype B in type 2 diabetes Acta Diabetol 2002 39 215 220 12486496 10.1007/s005920200037
Montani JP Carroll JF Dwyer TM Antic V Yang Z Dulloo AG Ectopic fat storage in heart, blood vessels and kidneys in the pathogenesis of cardiovascular diseases Int J Obes Relat Metab Disord 2004 28 Suppl 4 S58 S65 15592488 10.1038/sj.ijo.0802858
Lamarche B Tchernof A Mauriege P Cantin B Dagenais GR Lupien PJ Despres JP Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic heart disease JAMA 1998 279 1955 1961 9643858 10.1001/jama.279.24.1955
Sniderman AD Lamarche B Tilley J Seccombe D Frohlich J Hypertriglyceridemic HyperapoB in Type 2 Diabetes Diabetes Care 2002 25 579 582 11874951
Rainwater DL Mitchell BD Comuzzie AG Haffner SM Relationship of low-density lipoprotein particle size and measures of adiposity Int J Obes Relat Metab Disord 1999 23 180 189 10078854 10.1038/sj.ijo.0800813
Tanko LB Bagger YZ Alexandersen P Christiansen C Larsen PJ Peripheral adiposity and cardiovascular risk (letter) Circulation 2003 108 e164 e164 14662699 10.1161/01.CIR.0000102950.15230.95
Lemieux I Energy partitioning in gluteal-femoral fat: does the metabolic fate of triglycerides affect coronary heart disease risk? Arterioscler Thromb Vasc Biol 2004 24 795 797 15132969 10.1161/01.ATV.0000126485.80373.33
Kahn HS Why are large legs protective? (letter) Am J Clin Nutr 1997 66 712 713 9280198
Bengtsson C Bjorkelund C Lapidus L Lissner L Associations of serum lipid concentrations and obesity with mortality in women: 20 year follow up of participants in prospective population study in Gothenburg, Sweden BMJ 1993 307 1385 1388 8274890
Tanko LB Bagger YZ Qin G Alexandersen P Larsen PJ Christiansen C Enlarged waist combined with elevated triglycerides is a strong predictor of accelerated atherogenesis and related cardiovascular mortality in postmenopausal women Circulation 2005 111 1883 1890 15837940 10.1161/01.CIR.0000161801.65408.8D
Okura T Nakata Y Yamabuki K Tanaka K Regional body composition changes exhibit opposing effects on coronary heart disease risk factors Arterioscler Thromb Vasc Biol 2004 24 923 929 15016639 10.1161/01.ATV.0000125702.26272.f6
Di Loreto C Fanelli C Lucidi P Murdolo G De Cicco A Parlanti N Ranchelli A Fatone C Taglioni C Santeusanio F De Feo P Make your diabetic patients walk: Long-term impact of different amounts of physical activity on type 2 diabetes Diabetes Care 2005 28 1295 1302 15920042
Wang J Thornton JC Bari S Williamson B Gallagher D Heymsfield SB Horlick M Kotler D Laferrere B Mayer L Pi-Sunyer FX Pierson RNJ Comparisons of waist circumferences measured at 4 sites Am J Clin Nutr 2003 77 379 384 12540397
|
16150143
|
PMC1236917
|
CC BY
|
2021-01-04 16:30:07
|
no
|
BMC Cardiovasc Disord. 2005 Sep 8; 5:26
|
utf-8
|
BMC Cardiovasc Disord
| 2,005 |
10.1186/1471-2261-5-26
|
oa_comm
|
==== Front
BMC Cell BiolBMC Cell Biology1471-2121BioMed Central London 1471-2121-6-341615940210.1186/1471-2121-6-34Research ArticleEuploidy in somatic cells from R6/2 transgenic Huntington's disease mice Petersén Åsa [email protected]énius Ylva [email protected]örkqvist Maria [email protected] David [email protected] Neuronal Survival Unit, Department of Experimental Medical Science, Wallenberg Neuroscience Center, Lund University, Sweden2 Department of Clinical Genetics, University Hospital, Lund, Sweden3 Unit of Molecular Metabolism, Division of Diabetes, Metabolism, and Endocrinology, Department of Experimental Medical Science, Lund University, Sweden2005 13 9 2005 6 34 34 24 6 2005 13 9 2005 Copyright © 2005 Petersén et al; licensee BioMed Central Ltd.2005Petersén 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
Huntington's disease (HD) is a hereditary neurodegenerative disorder caused by a CAG repeat expansion in the HD gene. The huntingtin protein expressed from HD has an unknown function but is suggested to interact with proteins involved in the cell division machinery. The R6/2 transgenic mouse is the most widely used model to study HD. In R6/2 fibroblast cultures, a reduced mitotic index and high frequencies of multiple centrosomes and aneuploid cells have recently been reported. Aneuploidy is normally a feature closely connected to neoplastic disease. To further explore this unexpected aspect of HD, we studied cultures derived from 6- and 12-week-old R6/2 fibroblasts, skeletal muscle cells, and liver cells.
Results
Cytogenetic analyses revealed a high frequency of polyploid cells in cultures from both R6/2 and wild-type mice with the greatest proportions of polyploid cells in cultures derived from skeletal muscle cells of both genotypes. The presence of polyploid cells in skeletal muscle in vivo was confirmed by fluorescence in situ hybridisation with centromeric probes. Enlarged and supernumerary centrosomes were found in cultures from both R6/2 and wild-type mice. However, no aneuploid cells could be found in any of the tissues.
Conclusion
We conclude that polyploid cells are found in fibroblast and skeletal muscle cultures derived from both R6/2 and wild-type littermate mice and that aneuploidy is unlikely to be a hallmark of HD.
==== Body
Background
Huntington's disease (HD) is a hereditary neurodegenerative disorder caused by a CAG repeat expansion in the huntingtin gene [1]. It is characterized by personality changes, motor disturbances, cognitive decline, and weight loss [2]. Neuropathologically, the disease is reflected by neurodegeneration, primarily in the neostriatum, the cerebral cortex and the hypothalamus, with the appearance of cytoplasmic and intranuclear aggregates of misfolded huntingtin in neurons [3]. In HD patients and transgenic models of the disease, wild-type (wt) and mutant huntingtin are expressed in most tissues [4]. Neither the pathogenetic mechanism of HD nor the normal function of huntingtin is fully understood. Huntingtin interacts with key players of the mitotic machinery, such as the microtubuli and the centrosomes [5]. Interestingly, when huntingtin was identified in 1993 [1], one of its few known sequence motifs with established functions was the HEAT repeats [6]. These sequences are also found in proteins involved in mitotic progression and chromosomal dynamics [7]. The R6/2 transgenic mouse is the most widely used model to study HD. It expresses exon 1 of the human HD gene and displays several features of HD such as motor dysfunction, neuronal huntingtin aggregates, weight loss, and premature death at around 14 weeks of age [8]. In fibroblast cultures derived from R6/2 mice, a reduced mitotic index and high frequencies of cells with multiple centrosomes and aneuploidy have recently been reported [9]. Aneuploidy is characteristic for neoplastic cells, cells exposed to carcinogens, and cells from cancer-prone patients with hereditary chromosome instability syndromes. However, the phenomenon has rarely, if ever, been described in neurodegenerative disease in the absence of an increased risk for neoplasia. To further explore this unexpected aspect of HD, we have now performed extensive cytogenetic analyses of fibroblasts, skeletal muscle cells, and liver cells from 6- and 12-week-old R6/2 and wt littermate mice. We found no evidence of aneuploidy in R6/2 cells, but a suprisingly high frequency of polyploid cells in cultures derived from both R6/2 and wt littermate mice.
Results
Cytogenetic analysis
Fresh biopsies were taken from ear lobes, abdominal skeletal muscles, peritoneal fat, and liver from sacrificed six- and 12-week-old mice (Table 1). Cells from R6/2 and wt littemate mice grew equally well in culture up to five passages, after which the culture process was terminated. Analysis of metaphase peritoneal fat fibroblasts, skeletal muscle cells, and liver cells from 6-week-old R6/2 and wt littermate mice showed a normal diploid chromosome complement in the vast majority of cells and a tetraploid chromosome number in 3–32% of the cells. Cells taken from the abdominal skeletal muscle and the liver also showed 2–19% polyploid cells, with chromosome numbers up to the decaploid level (N = 400; Fig. 1a). These polyploid cells were present in biopsies both from R6/2 and wt mice. The highest frequencies were found in abdominal muscle biopsies from wt animals. After in vitro propagation up to four passages of muscle cells from one wt and two R6/2 mice, the frequency of polyploid cells had increased, compared to the first passage, from 8% to 45% in the wt culture, including cells with up to 600 chromosomes, whereas the frequency of such cells had decreased in one of the R6/2 cultures and was largely unchanged in the other. No more than four aneuploid cells were found in any of the cultures. These invariably showed loss of only one or two chromosomes compared to the diploid or tetraploid levels, or loss of up to four chromosomes compared to polyploid levels. In skeletal muscle cells, ear lobe fibroblasts, and peritoneal fat fibroblasts from 12-week-old animals, similar results were obtained. No more than four aneuploid cells (including hypodiploid, hypotetraploid, and hypopolyploid cells) were detected in any of the week-12 cultures.
Table 1 Cytogenetic data.
Mouse number Biopsy site Passage number Ploidy < 2n Ploidy = 2n Ploidy 2n-4n Ploidy = 4n Ploidy > 4n Total number of cells analyzed Polyploid cells (%)a
Wild-type week 6
30 AM 1 1 29 1 11 10 52 19
31 PF 1 2 46 0 11 0 59 0
31 AM 1 1 26 0 19 4 50 8.0
31 AM 4 0 3 0 3 5 11 45
Wild-type week 12
76 EL 1 2 46 0 4 0 52 0
76 EL 5 0 50 0 17 0 67 0
76 PF 1 0 42 0 7 1 50 2.0
76 AM 1 0 10 0 1 1 12 8.3
77 EL 1 0 60 0 4 0 64 0
77 EL 5 0 22 3 20 5 50 10
77 EL 1 0 46 0 5 1 52 19
77 AM 1 0 85 0 30 2 117 1.7
R6/2 week 6
38 PF 1 0 45 0 5 0 50 0
38 AM 1 0 39 0 12 5 56 8.9
38 AM 4 0 32 0 32 1 65 1.5
38 LC 1 0 24 2 17 7 50 14
39 AM 1 0 33 1 12 4 50 8.0
39 AM 4 0 25 0 30 3 58 5.2
R6/2 week 12
78 EL 1 1 50 0 4 0 55 0
78 EL 5 0 39 1 19 3 62 4.8
78 PF 1 0 53 0 3 0 56 0
78 AM 1 0 30 0 4 1 35 2.9
79 PF 1 0 70 1 7 0 78 0
79 AM 1 0 50 1 7 0 58 0
a Cells with a ploidy level above the tetraploid (4n). AM, abdominal muscle cells; PF, peritoneal fat fibroblasts; EL, ear lobe fibroblasts; LC, liver cells.
Figure 1 Cytogenetic and immunofluorescent analysis. G-banded polyploid metaphase cell from wt31 abdominal muscle (a). Multiple centrosomes (red) in cells from the peritoneal fat of R6/2 78 (b; green autofluorescence indicates cytoplasm) and from the abdominal muscle of wt77 (c; autofluorescence removed for clarity). Tripolar anaphase cell visualized by haematoxylin-eosin staining in a cell (passage 1) from the abdominal muscle of wt30 (d). Large nucleus (arrow) in a skeletal muscle fibre from wt77 (e). One nucleus with two chromocenters (arrow; green) and one with eight chromocenters (arrowhead) in a skeletal muscle fibre from wt76 (f).
Centrosome morphology and mitotic polarity
Centrosome detection was performed on cultured skeletal muscle cells, peritoneal fat fibroblasts, and ear lobe fibroblasts (passage 1 and also passage 5 of the ear lobe biopsies) from 12-week-old mice. This revealed abnormally enlarged centrosomes (> 2 centrioles) in a small population of cells in all cases (1–4%). A proportion of these cells (0.5–2% out of all cells) had three or four centrosomes, whereas the remaining cells had normal centrosome numbers (one or two; Fig 1b and 1c). There were no significant differences between wt and R6/2 cells (p > 0.05; > 100 cells analysed from each culture). Because supernumerary centrosomes have been associated with spindle multipolarity at mitosis, we also analysed cell division figures from the cultured biopsies of 6-week and 12-week mice after haematoxylin-eosin staining. Tripolar metaphase and anaphase configurations were detected in cells from the muscle biopsies from one of the 6-week old R6/2 mice and one of the 6-week old wt mice (2/160 cells and 2/232 cells respectively; Fig. 1d) but not in cells from any of the other biopsies (>150 cells analysed per biopsy).
Identification of polyploid cells in vivo
In both R6/2 and wt mice, the highest rate of polyploidy was noted in cultures established from the abdominal muscle. To corroborate these findings in vivo and to identify the cell type harbouring an abnormally high chromosome number, 10 μm tissue sections from two wt and two R6/2 mice were first examined after haematoxylin-eosin staining. This revealed a low frequency (1–2%) of nuclei with diameters three to four times the normal range residing in the striated muscle fibres (Fig. 1e). Such enlarged nuclei were neither found in the adjacent fibrous tissue, nor in the vascular tissue penetrating the muscles. To exclude that these nuclei were simply artifacts caused by oblique cutting of the sections, fluorescence in situ hybridization (FISH) with a murine pan-centromeric probe was performed on muscle tissue sections from one wt mouse. As expected, this revealed two chromocenters in the majority of nuclei (536/788 = 68%) and small populations of nuclei with one (10%), three (9%), or four (9%) chromocenters. These signal configurations were also found in the adjacent fibrous tissues. However, there was also a small population (4%) of large nuclei with seven or eight chromocenters (Fig. 1f), all of which were present in striated muscle fibres. Nuclei with > 4 chromocenters were also found in abdominal muscle fibres from the other 12-week-old wt mouse and the two R6/2 mice, in proportions of 1–4%.
Discussion
The function of normal huntingtin is not fully elucidated. Sequences of known functions within huntingtin include HEAT repeats, which are involved in chromosome dynamics [7]. In HD, the expanded polyglutamine in huntingtin is thought to alter its protein interactions [5]. A previous study described high frequencies of multiple centrosomes and aneuploidy in fibroblast lines derived from R6/2 mice and HD patients [9]. Aneuploidy is closely connected to neoplastic disease, as the vast majority of tumours exhibit acquired chromosome abnormalities [10]. Moreover, most inherited chromosome instability syndromes are associated with a significantly elevated risk of cancer. However, the risk of malignant tumours has been shown to be lower in HD patients than in the normal population [11]. To explore this ostensibly paradoxical situation, we performed cytogenetic analyses of cultured cells from multiple tissues in R6/2 mice. Although a high number of cells were analysed, we could find only minute populations of anueploid cells, occurring at similar frequencies in wt and R6/2 mice. These were typically hypodiploid, hypotetraploid, or hypopolyploid with loss of only few chromosomes, indicating that they were most probably due to artifactual loss of chromosomes during preparation. Hence, we found no convincing evidence for aneuploidy. The reason for the discrepancy of results between our study and the previous report [9] may arise from differences in the genetic background of the R6/2 colonies, different culture procedures, or differences in the methods for chromosome preparation. Furthermore, in our study, we defined aneuploidy according to the ISCN (1995) recommendations [12], whereas the previous study defined aneuploid cells as any cells harboring a chromosome number different than 2n = 40. Some of the cells scored as polyploid in our study might thus have been scored as "aneuploid" in the previous study by Sathasivam et al [9]. We found small populations of highly polyploid cells in liver, fibroblast and skeletal muscle cultures derived from R6/2 and wt littermate mice. In these cultures, derived from two mice per genotype, there was no difference in the frequency of polyploid cells between R6/2 and wt mice.
Polyploid cells have been found in several murine tissue including vascular smooth muscle [13], ovaries [14], cardiomyocytes [15], colonic fibroblasts [16], liver cells, bladder epithelial cells, uterine decidua cells, trophoblasts, and megakaryocytes [17]. Our study supports that polyploidy is a normally occurring phenomenon in many murine tissues and shows that skeletal muscle cells may have an unusually high frequency of polyploid cells. An earlier investigation describes a high frequency of multiple centrosomes in fibroblast cell lines from R6/2 mice and from HD patients [9]. In the present study, we found low frequencies of cells with multiple centrosomes in cultures from both R6/2 and wt littermate mice. In neoplastic cells, supernumerary centrosomes are associated with multi-polar cell divisions and have traditionally been suggested to generate aneuploid daughter cells [18]. Somewhat surprisingly, we found multipolar mitoses at a low frequency in skeletal muscle cultures from both genotypes. To our knowledge, this type of mitotic aberration has not been shown previously in non-neoplastic cells. Whether similar mitotic aberrations also occur in murine skeletal muscle in vivo remains to be established.
Conclusion
The present study shows that even though polyploidy may occur to a similar extent in wt and R6/2 cultured somatic cells, aneuploidy does not occur at a high frequency in R6/2 cells.
Methods
Transgenic mice
Transgenic HD mice of the R6/2 line were originally purchased from Jackson Laboratories (Bar Harbor, ME, USA) and the colony was maintained by breeding heterozygous R6/2 males with females from their background strain (F1 of CBA × C57Bl/6). Tails of the offspring were used to obtain DNA for determination of the genotype using a polymerase chain reaction assay [8]. The mice exhibit around 150 CAG repeats in the exon 1 of the HD gene. The mice were housed in groups with ad libitum access to food and water at a 12 h light/dark cycle and sacrificed at either 6 or 12 weeks of age using halothane anaesthesia. The experimental procedures were approved by the Ethical Committee at Lund University.
Cell culture and morphological analyses
Cells were cultured in RPMI 1640 medium, supplemented with 17% bovine serum, glutamine, and antibiotics. Chromosome preparation, G-banding, analysis of mitotic morphology, haematoxylin-eosin staining, and FISH were performed according to standard methods [19]. The number of metaphase cells analyzed from each biopsy, genotype, and age are specified in Table 1. Mouse major satellite sequences were detected in at least 500 interphase nuclei per biopsy using commercially available probes (Cambio, Cambridge, UK). Centrosome morphology was visualized by immuno-fluorescence using a monoclonal anti-γ-tubulin antibody (1:1000, GTU-88, Sigma, St. Louis, MS). At least 100 cells from each biopsy were analysed.
List of abbreviations
FISH fluorescence in situ hybridization
HD Huntington's disease
Wt wild-type
Authors' contributions
ÅP participated in the design of the study as well as the writing of the manuscript. YS carried out the cell culturing and the chromosome preparation. MB prepared the muscle tissue. DG performed the cytogenetic and FISH analyses and participated in the design of the study, and the writing of the manuscript.
Acknowledgements
We are grateful to Elsy Ling and Doris Persson for excellent technical assistance. ÅP and MB are supported by the Swedish Brain Foundation. YS and DG are supported by the Swedish Children's Cancer Foundation, the Sharon B. Lund Foundation of the American Cancer Society, and by the Donation Funds of the Lund University Hospital.
==== Refs
Huntington's Disease Collaborative Research Group A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes Cell 1993 72 971 983 8458085 10.1016/0092-8674(93)90585-E
Petersen Å Brundin P Huntington's disease: the mystery unfolds? Int Rev Neurobiol 2002 53 315 339 12512345
DiFiglia M Sapp E Chase KO Davies SW Bates GP Vonsattel JP Aronin N Aggregation of huntingtin in neuronal intranuclear inclusions and dystrophic neurites in brain Science 1997 277 1990 1993 9302293 10.1126/science.277.5334.1990
Landles C Bates GP Huntingtin and the molecular pathogenesis of Huntington's disease. Fourth in molecular medicine review series EMBO Rep 2004 10 958 963 15459747 10.1038/sj.embor.7400250
Li SH Li XJ Huntingtin-protein interactions and the pathogenesis of Huntington's disease Trends Genet 2004 20 146 154 15036808 10.1016/j.tig.2004.01.008
Andrade MA Bork P HEAT repeats in the Huntington's disease protein Nat Genet 1995 11 115 116 7550332 10.1038/ng1095-115
Neuwald AF Hirano T HEAT repeats associated with condensins, cohesins, and other complexes involved in chromosome-related functions Genome Res 2000 10 1445 1452 11042144 10.1101/gr.147400
Mangiarini L Sathasivam K Seller M Cozens B Harper A Hetherington C Lawton M Trottier Y Lehrach H Davies SW Bates GP Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice Cell 1996 87 493 506 8898202 10.1016/S0092-8674(00)81369-0
Sathasivam K Woodman B Mahal A Bertaux F Wanker EE Shima DT Bates GP Centrosome disorganization in fibroblast cultures derived from R6/2 Huntington's disease (HD) transgenic mice and HD patients Hum Mol Genet 2001 10 2425 2435 11689489 10.1093/hmg/10.21.2425
Mitelman F Johansson B Mertens F Mitelman Database of Chromosome Aberrations in Cancer 2005
Sorensen SA Fenger K Olsen JH Significantly lower incidence of cancer among patients with Huntington disease Cancer 1999 86 1342 1346 10506723 10.1002/(SICI)1097-0142(19991001)86:7<1342::AID-CNCR33>3.0.CO;2-3
Mitelman F ed An International System for Human Cytogenetic Nomenclature 1995 Basel: S. Karger
Jones MR Ravid K Vascular smooth muscle polyploidization as a biomarker for aging and its impact on differential gene expression J Biol Chem 2004 279 5306 5313 14634004 10.1074/jbc.M308406200
Keighren M West JD Analysis of cell ploidy in histological sections of mouse tissues by DNA-DNA in situ hybridization with digoxigenin-labelled probes Histochem J 1993 25 30 44 8432662 10.1007/BF00161042
Brodsky VY Delone GV Tsirekidze NN Genome multiplication in cardiomyocytes of fast- and slow-growing mice Cell Differ 1985 17 175 181 4053188 10.1016/0045-6039(85)90485-3
Neal JV Potten CS Polyploidy in the murine colonic pericryptal fibroblast sheath Cell Tissue Kinet 1981 14 527 536 7273094
Keighren MA Macfadyen LP Hill AS Patek CE Telfer EE West JD Polyploid cells in the mouse ovary J Anat 2003 202 563 571 12846477 10.1046/j.1469-7580.2003.00189.x
Hansemannn D Ueber patologische Mitosen Arch Pathol Anat Phys Klin Med 1891 119 299 326
Gisselsson D Jonson T Yu C Martins C Mandahl N Wiegant J Jin Y Mertens F Jin C Centrosomal abnormalities, multipolar mitoses, and chromosomal instability in head and neck tumours with dysfunctional telomeres Br J Cancer 2002 87 202 207 12107843 10.1038/sj.bjc.6600438
|
16159402
|
PMC1236918
|
CC BY
|
2021-01-04 16:31:30
|
no
|
BMC Cell Biol. 2005 Sep 13; 6:34
|
utf-8
|
BMC Cell Biol
| 2,005 |
10.1186/1471-2121-6-34
|
oa_comm
|
==== Front
BMC Fam PractBMC Family Practice1471-2296BioMed Central London 1471-2296-6-381616229010.1186/1471-2296-6-38Research ArticleAvoidance as a strategy of (not) coping: qualitative interviews with carers of Huntington's Disease patients Lowit Alison [email protected] Teijlingen Edwin R [email protected] Department of Mental Health, University of Aberdeen, IMS Building, Medical School, Aberdeen, AB25 2ZD, UK2 Department of Public Health & Dugald Baird Centre for Research on Women's Health, University of Aberdeen, Medical School, Aberdeen, AB25 2ZD, UK2005 14 9 2005 6 38 38 12 4 2005 14 9 2005 Copyright © 2005 Lowit and van Teijlingen; licensee BioMed Central Ltd.2005Lowit and van Teijlingen; 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
Since Huntington's Disease (HD) is a familial disease with an average onset in the mid-thirties, one might expect that spousal carers are concerned with providing care for off-spring who may turn out to be affected.
Methods
This study involved ten face-to-face interviews with carers of spouses affected by HD in Northeast Scotland. Carers were recruited through two channels: a genetic clinic and the Scottish Huntington's Association (SHA). Interviews were conducted in carers' own homes. A thematic analysis of the transcripts was conducted.
Results
Although carers did worry about their children, they did not envisage being involved in their care. Many avoided talking about the disease, both within and outwith their family; this may have greatly reduced the level of support provided by family members. Conversely, avoidance was often accompanied by symptom-spotting. For example, several people had given up driving, before they were incapable of doing so. The explanation appears to be that they avoided getting into situations in which HD may express itself.
Support meetings seem to be valued amongst patients with other serious diseases and their carers, however, although all participants had had contact with the SHA, only one regularly attended meetings. It was felt that seeing others with HD provided a constant reminder of the possible effect of HD on the wider family, which seemed to outweigh the benefit of attending. Overall, the analysis highlighted 'avoidance' as a key theme.
Conclusion
Many denied symptoms of HD in their spouses, pre-diagnosis. All had pretended at some point that it was not happening, through ignoring early signs and 'obvious' symptoms. Some partners had refused to go to the doctor until it was no longer possible to deny symptoms. Formal health and social care seemed to play a very small role compared to informal care arrangements.
==== Body
Background
Huntington's Disease (HD) is a terminal inherited progressive neurodegenerative disorder of the central nervous system, characterised by a variety of symptoms that affect the patient's physical and mental health. Folstein [1] describes the disorder as having a "triad of clinical features", because it has a range of motor, cognitive and psychiatric symptoms. Symptoms include personality change, movement disorder, cognitive impairment and mental health problems [1]. The average duration of the disease is 16 years [2], but it can vary greatly; a range of 4 – 38 years was documented in Folstein's study. [1] Symptoms can start at any age, but typically occur between the ages of 35–40, often after the affected individual has reproduced and possibly transmitted the faulty gene to his/her offspring. Inheritance is autosomal dominant and each child of an affected person has a 50% risk of inheriting the condition and developing HD. [2]
The gene causing HD (IT15) was isolated in 1993. Prior to this, the only possible method for presymptomatic HD testing was by linked DNA markers. This was a complex procedure that would not always give a conclusive result, and needed blood samples from multiple family members. The isolation of the single specific mutation meant testing became possible, with a reliability of >99%, without the need for co-operation of relatives. This has been of profound importance to HD family members, allowing a simple and accurate molecular test to determine their HD status. [2]
There are no treatments that can cure, delay onset or slow the course of HD; from onset onwards progressive degeneration occurs and the sufferer requires increasing levels of care and maintenance [2], but provision of a "full range of supportive medical nursing and social care" can help improve the patient's quality of life [28]. The complex characteristics and symptoms of HD mean caring for an affected individual can be particularly stressful and problematic. HD usually affects an individual at a time when they have many responsibilities to their family; this means the partner who acts as carer is often placed in the position of total responsibility, taking on the roles and responsibilities of their partner as well as their own. [3] HD places a great burden on the primary carer. Due to the hereditary nature of the disease, HD does not disappear with the death of the affected individual, but can repeat itself in successive generations. Consequently a carer may care for more than one generation of sufferers. [4]
Due to the rarity of HD many GPs (General Practitioners) and health workers do not have any real understanding of the disease, or the needs of patients and their carers. [5] An inadequacy of facilities for this population is very common, and management of the illness is very difficult. The complexity of the symptoms means that the HD patient is "unable to fit into the system". [6] Because HD patients' present symptoms that cross many different health and social care disciplines, there are often logistical and co-ordination problems associated with care provision. To be useful care has to be delivered timely, flexibly and in a co-ordinated fashion, to meet the patient's immediate requirements. All too often this does not happen, and the carer ends up being the central person in the day-to-day management of the patient.
This qualitative study was undertaken to gain in-depth understanding of the carer role in HD in the Northeast of Scotland; exploring the issues carers perceived were important when faced with the prospect of caring for one or more family members affected by HD. The study explored carers' perspective on their care-giving role and how that role had changed their lives. The themes explored were: motivations for caring and coping mechanisms adopted, plans for caring for future generations of HD relatives, utilisation of support services and facilitating factors and barriers to providing effective care.
Methods
The primary goal of qualitative research is to develop concepts, which help to understand social phenomena, emphasising the views, experiences and meanings of people experiencing the phenomena [7]. In order to answer the research question it was important to explore people's attitudes, views and experiences, therefore face-to-face interviews were selected as the most appropriate qualitative method. With the semi-structured interview there are no assumptions made or theories to prove; the method begins with an area of study and allows relevant data to emerge during the research process. [8] This type of interview aims to be like a guided conversation with a purpose. [9,10]
Ethical approval was applied for and was granted by the regional Medical Research Ethics Committee. Permission to approach clients was sought from the regional HD genetic clinic and the Scottish Huntington's Association (SHA). Due to the time scale of the study a convenience sample of ten carers was recruited from the Dundee, Angus and Grampian areas (the Northeast of Scotland).
Criteria for participant inclusion were that they (a) spoke English; (b) provided hands-on care for a family member who had been diagnosed with HD; and (c) were not participating in any other on-going research. A letter of invitation to the study, explaining the background to the research and the carer's potential role in the study was given to potential participants by a SHA adviser or the clinician at the regional HD genetics clinic, along with an acceptance slip and stamped, addressed envelope. Replies were sent to the first author at the university, who then arranged the time and place for the interview. Ten potential participants were approached and all agreed to be interviewed.
All the interviews were conducted in the carers' own home to make them feel more at ease. As many feel it is important to create a relaxed atmosphere in a comfortable setting to promote the development of trust. [9] Immediately before the interview the researcher explained the purpose of the study, the method of recording and analysing the data, and answered any questions the carers had, before obtaining their consent. When the study was completed participants were sent a summary of the findings.
Each interview started by asking the participants who they cared for, and what they did as a carer. This allowed participants to tell their own story and formed a starting point for the interview to develop. Themes were explored using open-ended questions. Questions were changed, dropped or expanded upon depending on each individual's experiences. The interviews were tape-recorded, [10] and transcribed verbatim within 24 hours, i.e. as soon as possible as the interviewer is more likely to recall exact details. [9]
The analysis consisted of reading and re-reading the transcripts to identify themes, patterns, salient points, common threads and trends, and to search for deviations and exceptions to these trends. [11,12] After the main themes were identified, each separate theme was explored individually to produce the final findings.
Several of the themes identified and explored in these interviews are beyond the scope of this paper (see Table 2). This paper focuses particularly on the theme of 'avoidance', and the part it plays in the coping strategies adopted by HD carers. Themes are presented by quotes from the interviews, which have been labelled with only an interview number to avoid possible identification of individual interviewees.
Table 1 Background data on carers/interviewees and patients
Carer Patient Relationship to Patient Children*
Male, (60–69) Female, (60–69) Husband 3
Male, (50–59) Female, (50–59) Husband 2-carer; 3-patient**
Male, (80–89) Female, (70–79) Husband 0
Female, (40–49) Male, (50–59) Wife 1 each**
Female, (50–59) Male, (50–59) Wife 3
Female, (40–49) Male, (40–49) Wife 3
Female, (60–69) Male, (60–69) Wife 4
Female, (50–59) Male, (50–59) Wife 2
Male, (50–59) Female, (40–49) Husband 2
Female, (40–49) Male, (40–49) Wife 0
* Number of children per couple have been slightly changed to maintain anonymity.
** Both have children from previous marriages only
Results
The ten carers were interviewed, were between 40 to 90 years old; six were female and four male. They were all spouses of the affected person. All but one carer cared for their spouse at home. The one carer, whose spouse was in a nursing home, had up until eight months ago, cared for her husband at home. The duration of the relationship between carers and spouses ranged from nine to 42 years. All stages of HD were represented in the sample. Table 1 summarises the sample.
The families in the sample had 24 children in total. Thirteen children were still at 50 percent risk of having the HD gene, a further eight had been tested for the HD gene; five are negative and three are positive for the faulty gene.
Only two carers knew about HD in their spouse's family, and the implications of the disease, before they were married. The partners in one couple had both been married before and knew the spouse had HD prior to the marriage. The other couple married not knowing the spouse's HD status, but decided not to have children.
One carer had been told that there was HD in the spouse's family, but had been told that there were no implications for them as a couple. One carer suspected HD in the spouse's family, although the family themselves were unaware of it. Six carers did not know about the existence of HD in the spouse's family prior to marriage. Three found out before their spouse's diagnosis and three found out later.
Table 2 Additional themes identified in interviews
Themes Examples
Motivation for being a carer • Marriage vows
• Memories of past life together
Changing roles and relationships • Demise of reciprocal loving relationship
• Loss of independence
Coping mechanisms • Living environment
• Slow course of HD
Employment & financial changes • Loss of employment
• Lower income
Experiences of social/ health care services • HD management
• Respite care
Scottish Huntington's Association • Practical advise
• Emotional support
Living with HD • Impact of the disease
• Personal development of carers
Impact of HD
HD has implications that extend far beyond the patient. One carer summarised the impact of HD on her spouse's family. Speaking about her father-in-law she said:
"He knows what's happening because his wife died with it. It's hard for him because they are his children. Three of his family have it and some of his grandchildren. It's not easy. (Husband's name removed) older sister has it and her son has it, though he doesn't have symptoms yet. His other sister has got it, but we don't know if any of her children have got it, they decided not to be tested. His younger brother we don't know about yet, because he's still quite young – in his thirties, and he hasn't been tested. He has children too. It's hard. It takes over everything, because quite a lot of his family are going to have to go through this, some are already." (Interview 4)
The initial diagnosis of HD was very difficult for those carers who had some knowledge of the disease. For some the shock lasted a long time:
"When we first found out he had it, it was hard. I was traumatised for quite a while afterwards." (Interview 6)
One carer, who had not been aware of HD, believed her ignorance had not made much difference to the way she handled her situation. She said:
"I don't know what I would have done – you may think I would have been better prepared – I might have been financially because we certainly were not – it was a nightmare, but nothing can prepare you for this. I don't believe there is anything that can prepare you for the horror of HD. I just wouldn't wish it on anyone. I worry mostly because of the children." (Interview 5)
Avoidance
Avoidance was a strong theme that was manifest in family members involved with HD. Avoidance often took the form of refusal to discuss HD within the family. One carer whose spouse had been diagnosed said:
"Well you know, he wouldn't talk about it at all. It was difficult to find anyone to talk to within his family. I didn't know about it (HD) and they weren't going to tell me anything about it." (Interview 5)
The five carers who knew that their spouse was at risk before their diagnosis also said that they avoided talking about or thinking about HD:
"We just got on with our lives without thinking about it. We never discussed it. What was the point? He could just as easily not have had it. A 50 percent risk is a 50 percent risk!" (Interview 10)
The main motivation behind refusal to discuss the issue in couples where risk is known seems to be a form of denial. Carers wanted to believe that their spouse was not affected by HD, and avoiding dealing with the risk factor reinforced this.
"I put it in a little box in the back of my mind because really I was hoping she wouldn't have it." (Interview 1)
For some, avoidance can be so strong that denial of obvious symptoms makes the diagnosis even more distressing, as the following interview highlights:
"After seeing her mother, well I had spent so long hoping she didn't have it and kind of ignoring her symptoms, that when I was told it kind of knocked me for six." (Interview 9)
After initial diagnosis, this inability to discuss HD continued for all the carers. For some, discussion seemed pointless in the face of inevitability, which of course, is not exactly the same as avoidance:
"I just think you have to keep going as normal as possible for as long as possible. Why discuss it – what's the point?" (Interview 8)
There was also a lot of secrecy surrounding HD within families. The secrecy was also present within the spouse's own family. This secrecy may have contributed to the inability to speak about the disease. One carer who believed that she had been kept in the dark about the implications of the disease was very angry and resentful:
"I was so angry – I felt angry with him and his dad. I suppose I wanted to protect my children. I wanted it to be all right for them. I was angry. He said he didn't know the implications of the disease, but I'm sure he and his dad knew more than they told me. Looking back I am sure of it." (Interview 5)
Another carer knew an hereditary disease existed in her partner's family, but had not been given any details about the disease. She suggested that if she had known, she might not have married her partner. Avoidance manifests itself at different levels – HD was not hidden from her, but no details about the disease were provided and this carer did not seek information for herself:
"I never met his dad, his mum never spoke about it, neither did he........ I knew when we married that this disease existed in his family, but I can honestly say I never thought about it. As I say, I had never met his dad, maybe if I had it would have been different." (Interview 10)
Although family secrecy was the cause of distress and anger for some carers, avoiding confronting the existence of the disease and discussing its implications sometimes led to a repetition of secrecy in the next generation:
"Well we never spoke about HD, never spoke about that, the kids never knew about it." (Interview 8)
Conversely, although all the carers talked about an unwillingness to talk about HD, several mentioned that they did look for the first signs of HD symptoms:
"I started noticing slight things about six years ago. I often wonder though, if I would have noticed them if I wasn't aware of the Huntington's." (Interview 4)
Symptom spotting and avoidance are conflicting behaviours:
"When I spotted the first signs I just kind of hoped that it wasn't it. I suppose I knew it was, but still I didn't think about it at the time." (Interview 9)
Recognising the symptoms, may have affected the way the spouses behaved. Some patients avoided initial diagnosis. One carer stated:
"It was a long time before she would go and see a doctor. A very long time, I had to persuade her to go because she knew herself what was coming and just didn't want to know. I'm sure it was that." (Interview 1)
Some spouses avoided situations where their failing abilities might be exposed. For example, four of the carers said that their spouses had given up driving, long before they had been diagnosed:
"He felt he was having problems driving and I was getting a bit worried about the driving situation...but he just gave up driving – I did all the driving." (Interview 10)
None of the couples indicated that they prepared for the long-term future or planned for the care of future generations of HD sufferers. Owing to the untreatable nature of HD, and its inevitable progression, avoidance and denial may have been part of coping mechanisms. Avoidance also influenced carers' ability to plan ahead. Their strategy was quite simple, they just dealt with problems as they arose:
" You can't really plan for it – I just take one day at a time. It's so slow you just adapt. I don't make plans." (Interview1)
Although the six couples who had children together all believed that the risk of transmission to their children was the worst part of the illness, they still avoided serious consideration of the long-term implications of HD:
"The implications for my children, that's the hardest thing. I find it so difficult to cope with. In fact I don't cope with it." (Interview 6)
The carers with children at risk were asked if they thought that they might be involved in the care of their own children. All but one carer said that it was a subject they never thought about, and they certainly didn't prepare for it. Typical responses were:
"No – I don't think about that (caring for child), not at all, never. It's not something you can think about." (Interview 8)
And
"I haven't thought about that – no. I hope they haven't got it. That's all I think about it." (Interview 9)
The one carer who had given the matter serious consideration had an affected daughter. However, for personal reasons she was not actively involved in caring for the daughter.
Although all the children knew of their risk, once the initial questions were answered all but one carer said that they never talked about it again:
"They know their dad is ill and it's going to get worse, and they know that they are probably at risk of getting it as well. And to me that's enough for them to be getting on with. We didn't talk about it because, for me, that would just pile it on." (Interview 8)
Avoidance was a strong theme here also:
"I don't really talk about it with them (children). We have spoken about it, but not now. Maybe I just hope they haven't got it." (Interview 9)
Issues around testing
All the carers with children at risk from HD said that they didn't put any pressure on them to be tested for the HD gene. All thought that this was a decision that their children should make for themselves. Of the children at risk less than half had undertaken the test. First hand knowledge of the symptoms and course of the disease acquired from close proximity to an affected parent, appears to influence the decision not to take the test. One carer, discussing or perhaps justifying why her husband had not been tested, explains:
"He lived through his father's illness from start to finish, so he didn't want to be tested. He said that he didn't want to know, it had been so awful that he wouldn't have been able to cope...knowing what was in store for him." (Interview 10)
There was also some concern that the genetic implications made the relationship between the children and the affected spouse distant. One carer explained:
"It has been difficult for them, seeing their dad change and what that might mean for them. They all find it difficult to be with him...it's not just the change in their father, it's partly that, but it is also that they see what might be in store for themselves." (Interview 7)
Problems arose both when a child had been tested and when it had not. One carer whose daughter had tested positive said:
"My eldest has been tested and is positive, but she doesn't want anyone to know. I wonder how she feels (daughter) knowing she has it. She won't talk about it. I think she has hidden it away." (Interview 6)
Lack of support for carers
The carers felt they had a close relationship with their children, and that relationship helped them emotionally and greatly enhanced their quality of life. However none of these children were actively involved in the care of their affected parent. One of the two offspring still living at home was an adult, but she did not help with any of the care of the affected parent, or upkeep of the house. This caused some friction within the home:
"We have a daughter who stays here. She doesn't lift a finger – she's 26, nearly 27, but she doesn't help either of us at all. I don't understand that, we argue about it all the time." (Interview 9)
None of the carers got any help from their spouse's parent. Several of these parents visited but none were involved in the day-to-day care. One carer tried to explain why that might be so – she stated:
"She had many problems looking after his father, especially as she had two children to bring up at the time. And really, I don't think she wanted to be involved in his care. She visits frequently, but doesn't help with his actual care." (Interview 10)
Another carer believed that it would be a great help if her spouse's mother could help her for a short time each week. She explained; this was not on offer:
"I think his mother should help me more. If she could take him out for a couple of hours a week that would really help. But I've asked her about that and she just shouted me down. She thinks it is my duty to look after him. She has no intention of giving me help." (Interview 6)
Several of the carers had no contact with their spouse's wider family and those that did said that it was just on a social basis that only happened once or twice a year.
SHA
This avoidance of contact with other HD families was also evident in the low attendance at support meetings. Although all the carers valued and praised the work of the HD voluntary service, all but one of the carers had stopped attending support meetings. They believed that, rather than helping them, meeting other carers made things worse, because they were presented with a vision of their future:
"Well I don't go. You kind of see what's in store for you, and it's not nice. Knowing – well, seeing what's going to happen. I found it too depressing."(Interview 7)
Regular attendance at SHA meetings did not fit in with the "one day at a time" attitude most carers had adopted:
"I went to one meeting, but they were all so miserable that I felt worse when I came out. I don't really want to know what's in store for me. I'd rather just cope with things as they happen." (Interview 10)
Discussion
The carers in this study had willingly accepted the responsibility of caring for an HD patient, and seemed to care with dedication and conscientiousness. The carers coped with the problems they encountered in day-to-day life due to their spouse's HD, by accepting the disease and trying to develop strategies to deal with the problems as they arose. The slow, progressive nature of HD facilitated this. This has been documented in studies exploring the caring role in other slow progressive diseases. [13-16] Many studies reporting on caring for family members with long-term progressive illnesses report that carers express a need to meet and share experiences with people facing the same type of situation. [13,17,18] However, the majority of carers in this study did not attend support meetings or wished to meet other HD carers.
The main reason given was that they did not like being presented with a vision of their own future. However, this does not adequately explain the contrast of attitudes between HD carers and carers of other progressive, degenerative diseases. Such carers are often active in attending support group meetings, where they too will be presented with a vision of their future. Given the hereditary nature of HD, the vision being presented may be more than simply the couple's own future, but also an unwelcome reminder of what may be in store for other family members.
Unlike other illnesses, which run a course and then disappear, albeit in death, HD affects whole families for generations. The hereditary nature of HD seems to impact on all aspects of life. Family breakdown frequently occurs, secrecy can be so great that it becomes impossible to discuss the subject within the family, and each member lives in isolation with his/her own feelings about the disease. [19] This may be because the scale of the problem witnessed in the sufferer is magnified by the genetic implications. Several studies found that many HD families adopt denial tactics to cope with the disease. Hiding the existence of HD within the family; not talking about HD, avoiding affected family members and denying one's own or one's spouse's symptoms will in some way avert the impending doom. [3,20] Refusal to allow information to be given to children has also been reported. [21,22] Evidence suggests that this inability to talk about HD is one of the reasons why the information about the disease held within families is so distorted, inaccurate and incomplete. [2,4,22]
Avoidance also affects support within the family. In this study none of the carers were getting any help with the care of their spouse from family members. Children, although close to the unaffected parent, were not involved in caring for their affected parent. The affected spouse's family also had no active involvement, and surprisingly the carer's family were not involved in helping with the affected spouse. Research has suggested that the burden of guilt and anxiety carried by members of HD families diminishes their ability to help and support affected individuals. [2,23]
Living with HD is essentially learning to live with stress, anxiety, fear and loss, which may help to understand the high level of avoidance and secrecy that existed within these families. Secrecy may result from parents' desire to protect children, which in turn may result in fragmentation of the family network as they avoid family members who have the disease, or are aware of the implications of the disease. Several of the carers had no contact with the spouse's wider family.
Once the initial diagnosis or "at risk" status was known and initial enquiries had been answered, HD was not mentioned inside or outside the family. The multiple problems which impinge upon all aspects of daily living, and the long-term prospects, may be too distressing to contemplate, so the full reality of the disease is hidden away and problems tackled only when they arise. However, this conscious avoidance is accompanied by symptom spotting and avoidance behaviour (such as giving up driving long before real symptoms arise), which suggests that although HD is not consciously spoken about, subconsciously it is ever present. This confirms previous reports on the effect of the all-consuming nature of HD on the affected families. [2,20,23,24]
By far the greatest problem for carers with children was the fear of their children developing the disease; all found the task of telling their children of their risk devastating. Once children had been informed of this, HD was not a topic that was openly discussed again. Protecting the children seems the main reason for this – speaking about HD would only make it more upsetting.
Only a small proportion of children had been tested for the faulty gene. This is consistent with the normal trend in HD families. [2,25,26] Before an accurate predictive test was available, continuous anxiety of asymptomatic family members lead to a belief that an accurate predictive test was urgently required and would be welcomed by HD families. However, the anticipated high uptake of the test has not been realised. [2] This may be due to asymptomatic individuals preferring to live with a 50 percent risk of having an incurable, degenerative disease, whose symptoms and course they are all too familiar with, rather than taking the test, which may conclusively indicate the presence of the disease.
Given the 50 percent risk each child and sibling of an affected individual has, the expectation was that carers might have considered being involved in care provision for other affected family members, and in anticipating future care they might have thought about how that care could best be provided. However, all the carers stated that this was not something they considered or prepared for. For them, this was something to deal with in the future, and certainly not a subject that they thought about or prepared for. The one carer who had an affected daughter stated that she would not be taking on the role of full-time carer for her daughter.
The study by Brouwer-Dudok de Wit and colleagues. which assessed the psychological distress of partners of people with late onset hereditary conditions revealed that partners of individuals at risk for HD had significantly higher levels of avoidance than partners of people at risk for hereditary cancers. [20] The ways in which individuals cope with stress has been extensively studied. [27] Avoidance is recognised as part of the coping process when individuals are confronted with threats they cannot deal with directly. [28] An individual judges a situation as traumatic and senses a painful effect such as depression or helplessness, so diverts his/her attention to something less painful. It is seen as a protective mechanism that allows us to proceed with life in greater psychological comfort. [27] Because HD is an incurable genetic disease combining mental and physical deterioration, the future may be too disturbing, and prognosis too hopeless, for carers to contemplate.
Avoidance and denial may be helpful in (temporarily) circumventing a serious problem but there is evidence that denying or suppressing painful thoughts can actually result in higher levels of anxiety and distress. [29] When a disease is serious, progressive and terminal, avoidance and denial may have serious consequences for the long-term well being of the carer. In such cases avoidance and denial may interfere with the maintenance of supportive relationships with partners, relatives and friends, and may preclude any form of long-term preparation. Several of the carers stated that they had no contact with their spouse's family; their children had very little input into the care of the affected parent and generally they felt distanced from any close supportive help with the day-to-day management of the condition. Skirton and Glendinning have argued that families could be supported and assisted in caring for their relatives in the home if appropriate. [30] We feel that avoidance and denial in some carers might prevent them from seeking and/or accepting such external support.
This study suggests that HD carers adopt an operational, or at best tactical approach to care, rather than a strategic one. Their considerations focus mainly on the immediacy of their current situation; they cope with their role by solving problems as they arise rather than planning coping strategies for future events.
Caring for an HD sufferer requires dedication and commitment, and it is a full-time occupation. At the same time the guilt and anxiety associated with hereditary disease [2,23] and the wide ranging financial and practical problems brought on by HD (not reported here) means avoiding thinking or talking about HD becomes an essential method of coping with day-to-day life. These combined factors mean that planning for future care of a relative with HD is probably beyond the carer's current capabilities or wherewithal.
This study is limited because it was based on a small sample. Future studies would benefit from accessing a larger, truly representative sample of HD carers. However, useful findings (for service provision as well as ideas for questions to be addressed in a large-scale survey) did emerge from the data.
Conclusion
Due to the wide range of symptoms of HD, carers face many different types of psychological, emotional and practical problems, which may need a wide range of health and social services. However, patients rely heavily on informal carers rather than formal services.
The genetic implications of HD, coupled with the lack of any cure or treatment to delay the course of the disease, means that avoidance is one of the main coping strategies deployed by carers, and secrecy about the disease is manifest in HD families.
In this study avoidance was found to impair carers' abilities to plan ahead or make anticipatory arrangements for the care of future relatives who may carry the faulty HD gene. Avoidance also curtailed effective participation in any network of support that might have come from other HD carers. Avoidance and secrecy were also found to hinder support from within the immediate and wider family.
List of abbreviations
HD Huntington's Disease
SHA Scottish Huntington's Association
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AL carried out the data collection and analysis for this study as part of her M.Sc. research. EvT supervised this research project and participated at all stages of the study. Both authors have written several drafts 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 interviewees without them this study would not have been possible. We would also like to thank Dr. Sheila Simpson and the Scottish Huntington's Association's Dundee and Angus advisers for their help in recruiting the study participants, as well as Karen Forrest Keenan for research advise.
==== Refs
Folstein SE Huntington's Disease: A Disorder of Families 1989 Baltimore: The John Hopkins University Press
Harper PS Huntington's Disease 1996 London: W. B. Saunders
Kessler S Forgotten person in the Huntington's Disease family Am J Med Genet 1993 48 145 150 8291568 10.1002/ajmg.1320480306
Korer J Fitzsimmons JS The effect of Huntington's Chorea on family life Br J Soc Work 1985 15 581 597
McGarva K Huntington's Disease – seldom seen, seldom heard? Health Bull (Edinb) 2001 59 306 308 12664744
Shakespeare J Huntington's Disease-falling through the net Health Trends 1993 25 19 23 10125696
Pope C Mays N Qualitative Research: Reaching the parts other methods cannot reach: an introduction to qualitative methods in health and health services research BMJ 1995 311 42 45 7613329
Flick U An Introduction to Qualitative Research 1998 London: Sage
van Teijlingen E Forrest K The range of qualitative research methods in family planning and reproductive health care J Fam Plann Reprod Health Care 2004 30 171 173 15222924 10.1783/1471189041261519
Rapley T Seale C, Gobo G, Gumbrium JF, Silverman D Interviews Qualitative Research Practice 2004 London: Sage 15 33
Forrest Keenan K van Teijlingen E Pitchforth E The analysis of qualitative research data in family planning and reproductive health care J Fam Plann Reprod Health Care 2005 31 40 43 15720849 10.1783/0000000052972825
Bowling A Research Methods in Health: Investigating Health and Health Services 1997 Buckingham: Open University Press
Pasetti C Rossi Ferrario S Fornara R Picco D Foglia C Galli J Caregiving and Parkinson's Disease Neurol Sci 2003 24 203 204 14598088 10.1007/s10072-003-0131-1
Boeije HR Duijnstee MSH Grypdonck MHF Continuation of caregiving amongst partners who give total care to spouses with multiple sclerosis Health Soc Care Community 2003 11 242 252 12823429 10.1046/j.1365-2524.2003.00422.x
Teijlingen van E Friend E Kamal AD Service use and needs of people with Motor Neurone Disease and their carers in Scotland Health Soc Care Community 2001 9 397 403 11846819 10.1046/j.1365-2524.2001.00320.x
Cheung J Hocking P Caring as worrying: the experience of spousal carers J Adv Nurs 2004 47 475 482 15312110 10.1111/j.1365-2648.2004.03126.x
Carers National Association Speak Up, Speak Out, Listen to Carers 1992 London: Issue Communications Ltd
Culley M The Nature Caring 1991 Aberdeen: Voluntary Services Aberdeen, Scotland
Bloch M Adam S Wiggin S Huggins M Hayden MR Predictive testing for Huntington's Disease in Canada: the experiences of those receiving a increased risk Am J Med Genet 1992 42 499 507 1535177 10.1002/ajmg.1320420416
Brouwer-Dudok de Wit ASC Savenije A Zoeteweij MW Maat-Kievit A Tibben A A hereditary disorder in the family and the family lifecycle: Huntington's Disease as a paradigm Fam Process 2002 41 677 691 12613124 10.1111/j.1545-5300.2002.00677.x
Yale R Huntington's Chorea – a family problem Br Med J (Clin Res Ed) 1981 282 1460 1461 6452928
Forrest K Simpson S Wilson BJ van Teijlingen ER McKee L Haites N Matthews E To tell or not to tell: barriers and facilitators in family communication about genetic risk Clin Genet 2003 64 317 326 12974737 10.1034/j.1399-0004.2003.00142.x
Hans MB Koeppen AH Huntington's Chorea, its impact on the spouse J Nerv Ment Dis 1980 168 209 214 6444999
Gilbert J Huntington's Disease – community care? A carer's tale J Public Health Med 1992 14 385 388 1489564
Tibben A Duivenvoorden HJ Vegter-van der Vlis M Niermeijer MF Frets PG van de Kamp JJP Roos RA Rooijmans HG Verhage F Presymptomatic DNA testing for Huntington disease: identifying the need for psychological intervention Am J Med Genet 1993 48 137 144 8291567 10.1002/ajmg.1320480305
Tibben A Frets PG van de Kamp JJP Niermeijer MF Vegter-van der Vlis M Roos RA van Ommen GJ Duivenvoorden HJ Verhage F Presymptomatic DNA-testing for Huntington disease: Pretest attitudes and expectations of applicants and their partners in the Dutch program Am J Med Genet 1993 48 10 16 8357031 10.1002/ajmg.1320480105
Oliver RW Psychology and Health Care 1993 London: Bailliere Tindall
Krohne HW Attention and Avoidance 1993 Seattle: Hogrefe & Huber
Wade SL Borawski EA Taylor HG Drotar D Yeates KO Stancin T The Relationship of Caregiver Coping to Family Outcomes during the Initial Year following Peadiatric Traumatic Injury J Consult Clin Psychol 2001 69 406 415 11495170 10.1037//0022-006X.69.3.406
Skirton H Glendinning N Using research to develop care for patients with Huntington's disease Br J Nurs 1997 6 83 90 9116444
|
16162290
|
PMC1236919
|
CC BY
|
2021-01-04 16:29:13
|
no
|
BMC Fam Pract. 2005 Sep 14; 6:38
|
utf-8
|
BMC Fam Pract
| 2,005 |
10.1186/1471-2296-6-38
|
oa_comm
|
==== Front
BMC GastroenterolBMC Gastroenterology1471-230XBioMed Central London 1471-230X-5-271609823410.1186/1471-230X-5-27Research ArticleAutoimmune hepatitis in India: profile of an uncommon disease Choudhuri Gourdas [email protected] Sanjay K [email protected] Chalamalasetty S [email protected] George [email protected] Department of Gastroenterology, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow 226014 (UP), India2005 15 8 2005 5 27 27 30 12 2004 15 8 2005 Copyright © 2005 Choudhuri 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
Autoimmune hepatitis (AIH) has been reported to show considerable geographical variation in frequency and clinical manifestations. It is considered a rare cause of liver disease in India. The present study was undertaken to determine the incidence, clinical, biochemical and histological profile of AIH in this part of the world.
Methods
Patients presenting with acute or chronic liver disease between January 1999 and June 2002 were evaluated prospectively. AIH was diagnosed using the international autoimmune hepatitis group criteria. Workup included clinical, biochemical, USG, viral markers, UGI endoscopy, AI markers (ANA, SMA, Anti-LKM, AMA, RF, p-ANCA) using indirect immunofluorescence and liver biopsy if possible.
Results
Forty-one of 2401 (1.70%) patients were diagnosed to have autoimmune liver disease. Out of these, 38 had autoimmune hepatitis and the rest 3 had primary biliary cirrhosis. The mean age of the patients of autoimmune hepatitis was 36.2 (15.9) years, 34 (89.4%) were females, and the duration of symptoms was 20.3 (20.5) months. Nineteen (50%) of them presented with chronic hepatitis, 13 (34.2%) as cirrhosis, 5 (13.1%) with acute hepatitis and 1 (2.6%) with cholestatic hepatitis. The presentations were jaundice in 21 (55.2%), pedal edema and hepatomegaly in 17 (44.7%), splenomegaly in 13 (34.2%), encephalopathy, abdominal pain in 9 (23.6%) and fever in 8 (21%). Twelve had esophageal varices and 3 had bled. Biochemical parameters were ALT 187 (360) U/L, AST 157 (193) U/L, ALP 246 (254) U/L, globulin 4.1 (1.6) g/dL, albumin 2.8 (0.9) g/dL, bilirubin 5.2 (7.4) mg/dL, prothrombin time 17 (7) sec and ESR 47 (17) sec. The autoimmune markers were SMA (24), ANA (15), both SMA and ANA (4), AMA (1), rheumatoid factor (2), pANCA (1), and Anti-LKM in none. Thirty (79%) patients had definite AIH and eight (21%) had probable AI hepatitis. Associated autoimmune diseases was seen in 15/38 (39.4%), diabetes 4, hypothyroidism 3, vitiligo 2, thrombocytopenia 2, rheumatoid arthritis 2, Sjogren's syndrome 1 and autoimmune polyglandular syndrome III in 1. Viral markers were positive in two patients, one presenting as acute hepatitis and HEV-IgM positive and another anti-HCV positive.
Conclusion
In India, autoimmune hepatitis is uncommon and usually presents with chronic hepatitis or cirrhosis, acute hepatitis being less common. Age at presentation was earlier but clinical parameters and associated autoimmune diseases were similar to that reported from the west. Primary biliary cirrhosis is rare. Type II AIH was not observed.
==== Body
Background
Autoimmune hepatitis (AIH) is a disease of unknown etiology characterized by chronic hepatocellular inflammation, serum autoantibodies, and hypergammaglobulinemia, which in most cases respond to immunosupression [1-3]. Those affected are mainly young women. The course is generally progressive and often fluctuating and cirrhosis is often present when the disease is discovered.
The diagnosis of AIH is established by the revised scoring system devised by the International Autoimmune Hepatitis Group and the international association for the study of liver [4,5]. The overall sensitivity of the score to establish a diagnosis of definite or probable AIH was 89.8%, however, the specificity for discriminating AIH from overlapping syndrome such as PSC or PBC was low [6].
Histological studies show periportal hepatitis with lymphocytic infiltrates, plasma cells, and piecemeal necrosis. Lobular hepatitis can be present. Presence of granulomas and iron deposition argues against AIH [7,8].
Autoimmune hepatitis affects 100,000–200,000 individuals in United States [9]. In India the prevalence is less [10-13]. Some early reports have questioned the existence of autoimmune liver disease. The prevalence, nature and prognosis of autoimmune hepatitis remain unclear.
In this report, we have studied the frequency, clinical, biochemical and immunoserologic profile of autoimmune liver disease. We also have compared the acute and chronic presentation of autoimmune liver disease.
Methods
Study population
Consecutive patients with chronic or acute liver disease seen at a tertiary care center Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India from January 1999 to June 2002 were evaluated for etiology.
Clinical assessment
In all patients, a detailed history was taken and clinical examination was carried out. History of onset of illness, acute or precipitating events, blood transfusion, surgery, menstrual abnormalities, and presence of extra-hepatic manifestations of autoimmune diseases were specifically noted. Family history of autoimmune diseases was also noted.
Laboratory tests and virological assessment
All patients underwent biochemical evaluation using standard automated techniques. Liver function tests and serum proteins and serum globulins were done in all patients. HBsAg, HBeAg, Anti-HBe were measured by immunoenzymatic assays (enzyme-linked immunosorbant assays, Hepanostika, Organon Technika, Boxtel). Anti-HCV was done using a commercially available qualitative ELISA against peptides corresponding to highly antigenic segments of core, NS3, NS4, NS5 regions of hepatitis C (UBI HCV EIA 4.0, Beijing United Biomedical Co., Ltd., China). Patients who were anti-HCV positive were further tested for HCV-RNA using a RT-PCR technique. IgM-anti HEV immunoenzymatic assays (enzyme-linked immunosorbant assays, Hepanostika, Organon Technika, Boxtel) were also done in patients presenting with acute hepatitis.
Immunoserologic assessment
Serological tests for autoantibodies antinuclear antibody (ANA), anti smooth muscle antibody (ASMA), anti-liver/kidney microsomal antibodies (anti-LKM-1) were done using the standard immunofluorescence technique. Briefly, cryostat sections of rat liver, kidney and stomach composite block were incubated with test sera (diluted 1:10) at 37°C for 30 min. After washing thrice with phosphate buffered saline (PBS) for 15 min each, the sections were incubated with fluoroscein isothiocyanate-conjugate rabbit anti-human polyclonal immunoglobulin (Dako, Copenhagen, Denmark) diluted 1:40 in PBS at 37°C for 30 min. The sections were washed thrice with PBS, mounted and examined under a fluorescent microscope. Titers of ANA, ASMA, and anti-LKM-1 of 1:80 or more is taken as positive in adults.
Histological assessment
Percutaneous liver biopsy was done wherever possible. An experienced histopathologist examined all the liver biopsy and specifically looked for lymphoplasmacytic infiltrate, piecemeal or bridging necrosis, rosette formation, bile duct injury, siderosis, copper deposits and granulomas. Fibrosis was graded as periportal, bridging or complete regenerative nodules (cirrhosis).
Upper gastrointestinal endoscopy for evaluation of varices and portal hypertensive gastropathy was done using a videoendoscope in all patients (Pentax gastroscope, EG 2940, Japan). Endoscopic retrograde cholangiography was performed in patients with alkaline phosphate was elevated more than three fold of upper limit of normal with a clinical suspicion of primary sclerosing cholangitis. Real time ultrasound of upper abdomen was done in all patients to look for evidence of portal hypertension, collaterals, and ascites and to rule out any obstructive biliary pathology. Appropriate tests to exclude Wilson's disease and hemochromatosis were done wherever indicated.
Diagnosis of autoimmune hepatitis was based on the criteria of the International Autoimmune Hepatitis Group criteria (Table 1).[5] Diagnosis of Overlap syndrome was made if the patient had clinical, serological and histological characteristics of two conditions either at the same time or during the course of their illness. None of the patients of AIH had history of significant alcohol consumption, blood transfusion in the preceding 3 years, homosexual contact or intravenous drug abuse.
Treatment
The indications for treatment were a greater than 2-fold elevation of aminotransferases in conjunction with interface hepatitis on liver biopsy. However in those in whom liver biopsy could not be done, elevation in liver enzymes alone was taken as indication for treatment. Response to treatment was defined as either or both of the following: marked improvement of symptoms and return of serum AST or ALT, bilirubin and immunoglobulin values completely to normal within 1 year and sustained for at least a further 6 months on maintenance therapy, or a liver biopsy showing at most minimal activity.
Statistical analysis
Results are presented as mean (S.D.) and range. SPSS 10.0 was used for statistical analysis. Data in the two groups (acute and chronic presentations) was compared using Mann-Whitney U test. P-value <0.05 was taken as significant.
Results
Of 2401 patients of liver disease who presented to the Gastroenterology department of Sanjay Gandhi Postgraduate of Medical Sciences, Lucknow, India, from January 1999 to June 2002, 41 (1.70%) patients fulfilled the International Autoimmune Hepatitis Group criteria [1]. Of these thirty-eight patients (1.5%) were confirmed to have autoimmune hepatitis and the other 3 had primary biliary cirrhosis based on liver biopsy findings. Thirty patients had definite (pretreatment score > 15) and eight had probable autoimmune hepatitis (pretreatment score 10–15) prior to treatment. But after initiation and continuation of treatment, 3 more attained scores of 17 and so were added to the definite autoimmune hepatitis group from the probable group. Among patients with autoimmune hepatitis, 35 (92.1%) were of type 1, none of type 2 and the rest 3 could not be classified as they lacked the SMA and ANA. The other etiologies of liver disease were hepatitis B in 30 %, hepatitis C in 20 %, alcohol in 25 %, cryptogenic in 15 % and others including autoimmune liver disease in 10 %.
Mean age of patients was 36.3 (2.6); range 7–68 years. Seven patients were less than 20 years and only five above 60 years of age. The age distribution at presentation showed a peak at 30 years (figure 1). There were 34 (89.4%) females; mean age of females [36.1 (16.7) range 7–68 years] was comparable with that of males [36.7 (13.8); range 11–51 years]. Mean duration of symptoms prior to presentation was 20.3 (3.9); range 0.2–72 months. Fifteen patients had duration of symptoms less than 6 months whereas 23 patients had chronic symptoms with duration of more than 6 months.
Clinical findings
Nineteen (50%) patients presented with chronic hepatitis, thirteen (34.2%) with cirrhosis, five (13.1%) with acute hepatitis and one (2.6%) with cholestatic hepatitis. Jaundice, edema and fatigue were the most common clinical presentations. Jaundice was usually mild only. Physical findings showed ascites in more than one-third of patients. Table 2 shows the clinical parameters.
Laboratory parameters
Mean alanine aminotransferase (ALT) was 187 (360) U/L and aspartate aminotransferase (AST) was 157 (193) U/L. Nine (22%) patients had AST and 11 (26.8%) had ALT above five times the upper limit of normal. The mean bilirubin was 5.2 mg/dl and nine patients (23.6%) had bilirubin above 5 mg/dL. See table 3 for the laboratory parameters.
Upper gastrointestinal endoscopy
Esophageal varices were present in twelve (31.5%) patients of whom three had presented with bleed. Portal hypertensive gastropathy was present in three patients and gastric varices in two patients.
Histological features
Liver biopsy was possible in only 19 (50%) patients, as 9 patients had coagulopathy, 6 had ascites and 4 had both ascites and coagulopathy, which contraindicated percutaneous liver biopsy. Chronic hepatitis was present in 15 and cirrhosis in 4. Interface hepatitis was present in 72.7%, rosette formation in 18.2% and lymphoplasmacytic infiltrates in 63.6%. Bile duct injury was observed in three patients.
Virological markers
In the one patient tested positive for anti-HCV, the serum also tested positive for HCV-RNA by PCR. One other patient presented as acute hepatitis and tested positive for IgM-HEV. Patient improved clinically but continues to be SMA positive and has elevated transaminases. Acute hepatitis E led to the recognition of the underlying asymptomatic chronic autoimmune hepatitis. One patient had HIV ELISA positive and had chronic diarrhoea with partial villous atrophy on duodenal biopsy.
Immunoserologic features
Antinuclear antibodies (ANA) was positive in 15 patients, of these 7 had speckled pattern (numerous evenly distributed specks of fluorescence, whereas remaining 8 had homogenous or diffuse pattern (uniform staining of nucleus). Anti-smooth muscle antibodies (SMA) were positive in 24 patients and out of these both ANA and SMA were positive in 4 patients. One of these patients had in addition AMA positivity and a high alkaline phophatase. But her liver biopsy did not reveal changes of PBC but showed features of autoimmune hepatitis (interface hepatitis and rosettes) and so she was classified under overlap syndrome (AIH-PBC). She had a score of 13, so falling under probable AIH. In 3 patients, no diagnostic antibodies were found and they were diagnosed on basis of other criteria including liver biopsy and response to immunosuppressive treatment. None of the patients had anti-liver/kidney microsomal (anti-LKM) antibodies.
Associated autoimmune diseases
Fifteen (39.4%) patients had associated autoimmune disease (Table 5). Diabetes mellitus, autoimmune thyroiditis and vitiligo were the most common extra hepatic autoimmune features.
Treatment outcomes
Thirty patients were started on treatment with corticosteroids (prednisolone) to which azathioprine was added in thirteen. In the remaining 8 patients, treatement with steroids was not started in 6 because of normal aminotransferases, presence of decompensated liver disease and liver biopsy (if done) not showing interface hepatitis and in another 2 because of being lost to follow up. Of the thirty patients started on steroids, 6 patients were lost to follow up and of the remaining 24, clinical and biochemical response was noticed in 17 (70.8 %). The patient with Overlap syndrome was treated with steroids but she did not achieve normalization of enzymes. Ursodeoxycholic acid was started in 5 patients in whom steroids either was not started or not tolerated, but only one patient achieved response. One patient on azathioprine had severe adverse effects in the form of pancytopenia and fever, requiring discontinuation of therapy. One patient had pulmonary tuberculosis in addition. He was on antitubercular drugs (isoniazid, rifampicin), on which he developed acute liver failure with hepatic encephalopathy and died due to increased intracranial tension. This patient was SMA positive. There were no other deaths in the study. One patient with decompensated cirrhosis who did not improve with immunosuppressant therapy underwent a liver transplant successfully and was doing well at 6 months follow-up.
Comparison of acute and chronic presentations did not show any significant difference between clinical, laboratory and immunoserologic parameters (Table 6). Histological parameters and treatment outcome also did not show any significant difference between these two groups.
Discussion
The overall prevalence of autoimmune liver disease in our study was 1.70% and of autoimmune hepatitis, 1.50%. This contrasts with data from Western studies where the estimated prevalence is 11–20% of all cases of chronic liver disease [3,17]. This difference with the Western data has also been seen in other Indian studies, which have reported a low prevalence of 3.5–6.1% of all cases of chronic liver disease [10,13]. The low prevalence in our population may be attributed to possible genetic or geographic variation. Studies looking at genetic predisposing factors have been largely directed at genes of the immunoglobulin superfamily, which include those encoding HLA located in the MHC, immunoglobulins and T-cell receptor molecules. Geographic variations in these factors are seen, with type 1 AIH in Caucasians being associated with HLA-DR3 serotype while in Japan where HLA-DR3 is rare, the primary association is with HLA-DR4. Similar differences in these as well as other multiple genes might account for some of this low frequency of AIH seen in India and has to be studied.
Majority of the patients were in the third and fourth decades with a mean age of 36.3 (12.6) years and there was a female preponderance (F: M 4:1) seen. Although it has long been appreciated that AIH more affects girls and young and middle-aged women, it has been recognized that the disease is not uncommon in the elderly. In a series of patients with type 1 AIH in northern European Caucasians, elderly patients were not uncommon and the mean age of patients from a study from Japan was 50.8 (12.7) years [1,14]. In the present study only 5 patients (13.1%) were above 60 years of age and this observation of onset at a younger age in Indian patients has been seen in other Indian studies in which the reported mean ages have been 31.0 (17.1) and 39.8 (13) years with a female predominance [10,13]. The mean duration of symptoms at diagnosis in our study was 20.3, which was longer than that reported form other countries as well as from India [10].
Most of our patients had advanced liver disease at presentation. Cirrhosis was present in 34.2% of patients and chronic hepatitis in 50% patients. The onset of autoimmune hepatitis is usually insidious, as described in western studies with fatigue, fluctuating jaundice and arthralgia as typical features, but a substantial proportion of patients have no obvious signs or symptoms of liver disease or have an acute presentation as seen in 25 % [18]. In our study presentation as acute hepatitis accounted for only 13.1 % of all presentations. In patients with advanced disease liver biopsy was not possible in most cases due to complications like coagulopathy and ascites. The relative higher prevalence of chronic hepatitis may be misleading, as the true prevalence of cirrhosis could be higher and might be picked up only on liver biopsy.
In general, the elevations of aminotransferases are more striking than those of bilirubin and alkaline phosphatase. In some cases of AIH, however, a cholestatic picture is present, marked by high levels of conjugated bilirubin and alkaline phosphatase. In our study, mean bilirubin concentration was only 5.2 mg% and only 9 patients had bilirubin above 5 mg %. One patient presented with high levels of bilirubin and alkaline phosphatase with ANA, SMA and AMA positvity. Her liver biopsy was typical of AIH and so she was classified under cholestatic autoimmune hepatitis.
Type 1 autoimmune hepatitis was present in 92.1% patients and type 2 was not seen. Other Indian reports also show higher prevalence of type 1 AIH, 88.9% from Mumbai and 80% from Delhi [13,10]. It has been described that about 70 to 80 % patients of AIH present with significant titers of ANA or SMA (or both) and about 3 to 4 % have anti-LKM-1 antibodies, while up to 20 % have none of these antibodies [4]. In our study as previously described 92.1 % had SMA or ANA (or both) and in the rest no antibody could be identified. Diagnoses in these latter group of patients were made on basis of other criteria such as marked hypergammaglobulinemia, typical histological findings, immunogenetic background, family history of autoimmune diseases, appropriate investigations to exclude other possible causes of liver disease and response to immunosuppressive treatment. There is currently no agreement on what constitutes an autoimmune overlap syndrome. Two distinct types of overlap syndrome can be considered [20]. The crossover syndrome in which an individual may fit one diagnosis while having some features associated with another and the true overlap in which the patient has clinical, serologic and histologic characteristics of two conditions either at the same time or during the course of their illness. Only one patient in this study had the true overlap syndrome of AIH-PBC as she had cholestatic hepatitis with positivity of ANA as well as AMA, and liver biopsy features of AIH. In the other Indian study by Gupta et al also, a low frequency of overlap syndrome was seen in two patients [10].
Concurrent immunological diseases are reported in 17 to 48% of patients with autoimmune thyroiditis, synovitis and ulcerative colitis being most common [1-3,16]. We found it in 39.4%, with diabetes, thyroiditis and vitiligo being the most common. One patient had autoimmune polyglandular syndrome. Anti-HCV was positive in only one patient (2.8%) in whom the HCV-RNA was negative by PCR. A study from Japan, where hepatitis C has high prevalence, reported anti-HCV positivity in 12.5% of patients of autoimmune hepatitis [14]. Approximately 5 % of patients with chronic hepatitis C have ANA or SMA titers of 100 or higher and so in those patients where anti-HCV is positive, HCV-RNA has to be done also to rule out hepatitis C virus infection as the cause of the autoimmune phenomena [19].
Steroid monotherapy or combination therapy with prednisolone and azathioprine are the standard initial treatment of AIH. Combination therapy is best suited for elderly, osteoporotic patients, those with diabetes, hypertension, obesity and psychiatric disorders. Monotherapy with steroids is preferred in patients with hematological abnormalities and in young patients with fertility concerns. Ursodeoxycholic acid is a hydrophilic bile acid with immunomodulatary capability. Small-uncontrolled trials have shown clinical and biochemical improvement and a reduction in histological abnormality when given over 2 years [15]. Thirty of our patients received either prednisolone or azathioprine. Disease remission (clinical and biochemical) was achieved in seventeen patients. On the basis of limited data it has been shown that overlap syndromes of PBC or PSC and AIH with significant interface hepatitis may respond, at least in part to corticosteroids and such patients should receive a trial of immunosuppression [20]. The lone patient of overlap syndrome in the present study did not respond to a trial of steroids.
Comparison of acute and chronic presentation did not show any significant difference between clinical, laboratory, immunoserologic parameters, histology and treatment outcome. This indicates that autoimmune hepatitis of acute onset is in fact a chronic disease with abrupt onset of symptoms or acute exacerbation. The perceived acuteness or chronicity of the disease is a reflection of detection bias. Due to a long duration of sub clinical disease, it may be irrational to distinguish acute and chronic autoimmune hepatitis as both the presentations have similar clinical, biochemical, immunoserologic and histologic features.
Conclusion
In summary, our patients presented at an early age, had a longer duration of symptoms. Female preponderance was observed. Type 1 autoimmune hepatitis was the most common, whereas type 2 was not observed in our patients. Clinical presentation was with advanced liver disease (chronic hepatitis and cirrhosis) and acute hepatitis was less common. Most patients had SMA or ANA positive. Associated autoimmune diseases were common. Hepatitis C infection was uncommon in our patients. Comparison of acute and chronic presentations did not reveal any significant difference, questioning the need for any such distinction.
Abbreviations
ANA: anti-neutrophilic antibody
AMA: anti-mitochondrial antibody
Anti-LKM: anti-liver/kidney microsomal antibodies
p-ANCA: anti-neutrophil cytoplasmic antibody
RF: rheumatoid factor
PBC: primary biliary cirrhosis
PSC: primary sclerosing cholangitis
Competing interests
The author(s) declare that they have no competing interests.
Contributions
GC: conceived the study, participated in its design, coordination and drafting the manuscript
SKS, GA: participated in the study design, collection of data and drafting the manuscript
CSB, TSN: participated in the study design and collection of data
Pre-publication history
The pre-publication history for this paper can be accessed here:
Figures and Tables
Figure 1 Age distribution of patients (n = 38)
Table 1 International diagnostic criteria for diagnosis of AIH
Parameter Score
Gender
Female +2
Male 0
Serum biochemistry
Ratio of elevation of serum alkaline phosphatase vs. aminotransferase
>3.0 -2
1.5–3 0
<1.5 +2
Total serum globulin, γ-globulin, or IgG (times upper limit normal)
>2.0 +3
1.5–2.0 +2
1.0–1.5 +1
<1.0 0
Autoantibodies (titers by immunofluorescence on rodent tissue)
ANA, SMA, or LKM-1
>1:80 +3
1:80 +2
1:40 +1
<1:40 0
Antimitochondrial antibodies
Positive -4
Negative 0
Hepatitis viral markers
Negative +3
Positive -3
Other etiological factors
History of drug usage
Yes -4
No +1
Alcohol (average consumption)
<25 g/day +2
>60 g/day -2
Genetic factors: HLA DR3 or DR4 +1
Other autoimmune diseases +2
Response to therapy
Complete +2
Relapse +3
Liver histology
Interface hepatitis +3
Predominant lymphoplasmacytic infiltrate +1
Rosetting of liver +1
None of above -5
Biliary changes -3
Other changes -3
Seropositivity for other defined autoantibodies +2
Definite AIH, >15 before treatment and >17 after treatment; probable AIH 10–15 before treatment and 12–17 after treatment
Table 2 Clinical features (n = 38)
Clinical features No (%)
Symptoms
Jaundice 21 (55.2)
Edema 17 (44.7)
Fatigue 17 (44.7)
Encephalopathy 9 (23.6)
Pruritus 9 (23.6)
Abdominal pain 9 (23.6)
Fever 8 (21.0)
Arthralgia 7 (18.4)
Menstrual abnormalities 4 (10.5)
Physical signs
Hepatomegaly 17 (44.7)
Splenomegaly 13 (34.2)
Ascites 13 (34.2)
Vitiligo 3 (7.8)
Table 3 Lab parameters (n = 38)
Parameters Mean ± SD Range % with abnormal value
Bilirubin (mg/dl) 5.2 ± 7.4 (0.21–33.4) 63.9
ALT (U/L) 187 ± 360 (10–2100) 75.0
AST (U/L) 157 ± 193 (10–986) 77.8
ALP (U/L) 230 ± 230 (63–935) 41.7
Albumin (g/dL) 2.8+0.9 (1.2–4.3) 75.0
Globulin (g/dL) 4.6+1.5 (2.0–8.1) 58.3
ESR (sec) 45+18 (12–72) 55.0
Prothrombin time (sec) 17+7 (8–36) 50.0
Table 4 Autoimmune markers (n = 38)
Autoimmune marker n %
SMA 24 (63.1)
ANA 15 (39.4)
SMA + ANA 4 (10.5)
AMA 1 (2.6)
p-ANCA 1 (2.6)
Rheumatoid factor 2 (5.2)
Table 5 Extra hepatic manifestations
Associated autoimmune diseases n
Diabetes 4
Thyroiditis 3
Vitiligo 2
Thrombocytopenia 2
Rheumatoid arthritis 2
Sjogren's syndrome 1
AI polyglandular syndrome type III 1
Table 6 Initial findings in autoimmune hepatitis with acute and chronic presentations
Initial findings Acute presentation (n = 15) Chronic presentation (n = 23)
Clinical
Age (years) 36.6 (16.5) 36.0 (16.0)
Sex (F: M) 13:3 4:1
Laboratory parameters
ALT (10–40 U/L) 143 (139) 213 (457)
AST (10–40 U/L) 187 (247) 207 (393)
ALP (U/L) 209 (174) 251 (280)
Bilirubin (0.2–0.8 mg/dL) 7.3 (9.6) 3.6 (4.7)
Albumin (3.5–4.5 g/dL) 2.6 (0.9) 3.1 (0.8)
Globulin (2.5–3.5 g/dL) 4.8 (1.8) 4.5 (1.4)
Prothrombin time (sec)* 19 (8.2) 15.9 (4.5)
Immunoserologic markers
ANA n (%) 6 (40.0%) 9 (39.1%)
SMA n (%) 9 (60.0%) 15(65.2%)
AMA n (%) 0 1 (4.3%)
Values expressed as Mean (S.D.)
* Control 12.7 sec
Comparison of parameters in the two groups did not show any statistical difference.
==== Refs
Czaja AJ Autoimmune hepatitis: Evolving concepts and treatment strategies Dig Dis Sci 1995 40 435 438 7851212 10.1007/BF02065434
Czaja AJ Krawitt EL, Weisner RH Diagnosis, prognosis, and treatment of classical autoimmune chronic active hepatitis Autoimmune liver disease 1991 New York, Raven Press 143
Czaja AJ Natural history, clinical features, and treatment of autoimmune hepatitis Semin Liver Dis 1984 4 1 6710168
Johnson PJ McFarlane IG Meeting report: International autoimmune hepatitis group Hepatology 1993 18 998 1005 8406375
Alvarez F Berg PA Bianchi FB Bianchi L Burroughs AK Cancado EL Chapman RW Cooksley WG Czaja AJ Desmet VJ Donaldson PT Eddleston AL Fainboim L Heathcote J Homberg JC Hoofangale HH Kajumu S Krawitt EL Mackay IR MacSween RN Maddrey WC Manns MP McFarlane IG Meyer zum Buschenfelde KH Zeniya M International autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis J Hepatol 1999 31 929 938 10580593 10.1016/S0168-8278(99)80297-9
Manns MP Strassburg CP Autoimmune hepatitis: Clinical challenges Gastroenterology 2001 120 1502 1517 11313321 10.1053/gast.2001.24227
Manns MP Strassburg CP O'Grady JG, Lake JR, Howdle DP Autoimmune hepatitis Comprehensive clinical Hepatology 2000 16 London: Mosby 1 14
Dienes HP Popper H Manns M Baumann W Thoenes W Meyer zum Buschenfelde KH Histologic features in autoimmune hepatitis Z Gastroenterol 1989 27 325 330 2505455
Jacobson DL Gange SJ Rose NR Epidemiology and estimated population burden of selected autoimmune diseases in United States Clin Immunol Immunopathol 1997 84 223 245 9281381 10.1006/clin.1997.4412
Gupta R Agarwal SR Jain M Malhotra V Sarin SK Autoimmune hepatitis in the Indian subcontinent: 7 years experience J Gastroenterol Hepatol 2001 16 1144 1148 11686842 10.1046/j.1440-1746.2001.02602.x
Agarwal A Singh RR Misra R Chronic active hepatitis with multisystem disease: Lupoid hepatitis or systemic lupus erythematosus? J Ind Rheum Assoc 1995 3 65 66
Balakrishnan C Mangat G Kalke S Desai D Joshi A Deshpande RB Joshi VR The spectrum of chronic autoimmune hepatitis J Assoc Physicians Iindia 1998 46 431 435
Amrapurar DN Amrapurkar AD Role of autoimmunity in nonviral chronic liver disease JAPI 2000 48 1064 1069 11310382
Toda G Zeniya M Wantanabe F Imawari M Kiyosawa K Nishioka M Tsuji T Omata M the Japanese national Study Group of Autoimmune Hepatitis Present status of autoimmune hepatitis in Japan- correlating the characteristics with international criteria in an area with a high rate of HCV infection J Hepatol 1997 26 1207 1212 9210605 10.1016/S0168-8278(97)80453-9
Nakamura K Yoneda M Yokohama S Tamori K Sato Y Aso K Aoshima M Hasegawa T Makino I Efficacy of Ursodeoxycholic acid in Japanese patients with type 1 autoimmune hepatitis J Gastroenterol Hepatol 1998 13 490 495 9641646
Czaja AJ Carpenter HA Santrach PJ Evidence against hepatitis viruses as important cause of severe autoimmune hepatitis in United States J Hepatol 1993 18 342 8228128 10.1016/S0168-8278(05)80279-X
Czaja AJ Diagnosis and therapy of autoimmune liver disease Med Clin North Am 1996 80 973 94 8804371
Nikias GA Batts KP Czaja AJ The nature and prognostic implication of autoimmune hepatitis with acute presentation J Hepatol 1994 19 1513 1520 10.1016/0270-9139(94)90250-X
Zauli D Cassani F Muratori L Krawitt EL, Wiesner RS, Nishioka M Autoantibodies in hepatitis C Autoimmune liver diseases 1998 2 Amsterdam: Elsevier Science 331 341
Woodward J Neuberger J Autoimmune Overlap Syndromes Hepatology 2001 33 994 1001 11283866 10.1053/jhep.2001.23316
|
16098234
|
PMC1236920
|
CC BY
|
2021-01-04 16:03:27
|
no
|
BMC Gastroenterol. 2005 Aug 15; 5:27
|
utf-8
|
BMC Gastroenterol
| 2,005 |
10.1186/1471-230X-5-27
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1111612902510.1186/1471-2164-6-111Research ArticleHypoxia-activated genes from early placenta are elevated in Preeclampsia, but not in Intra-Uterine Growth Retardation Vaiman Daniel [email protected] Françoise [email protected]ès-Duran Alexandra [email protected] Thérèse-Marie [email protected] Brigitte [email protected] Régis [email protected] Hélène [email protected] Sonia T [email protected] Frédérique [email protected] Geoffrey [email protected] Vincent [email protected] François [email protected] Jean-Louis [email protected] Virginie [email protected] Bruno [email protected]é Françoise [email protected] Génétique et Epigénétique des Pathologies Placentaires, GEPP, U709 INSERM-Université René Descartes-Institut Alfred Jost, Pavillon Baudelocque, Hôpital Cochin, 123 Boulevard de Port-Royal, 75014, Paris, France2 Laboratoire de Biochimie Medicale, Faculte de Medecine et de Pharmacie, UMR INSERM U.384 UA, 28 Place Henri Dunant, BP. 38, 63000 Clermont-Ferrand, France3 Centre de Ressources Biologiques, Laboratoire de Radiobiologie et d'Etudes des Génomes, Centre de Recherches INRA de Jouy-en-Josas, INRA, CRJJ, 78352 Jouy-en-Josas, France4 UPR CNRS 9078, Université René Descartes ParisV, Site Necker, 156 rue de Vaugirard, 75015 Paris, France5 Service de Réanimation Néonatale, Institut de Puériculture et de Périnatalogie, 26, boulevard Brune, 75014 Paris, France6 Service de Gynecologie-Obstetrique, Hopital Saint Antoine, 184 rue du Faubourg Saint Antoine, 75012 Paris, France7 Département de Génétique Animale, INRA, 78352, Jouy-en-Josas,, France2005 29 8 2005 6 111 111 10 5 2005 29 8 2005 Copyright © 2005 Vaiman et al; licensee BioMed Central Ltd.2005Vaiman 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
As a first step to explore the possible relationships existing between the effects of low oxygen pressure in the first trimester placenta and placental pathologies developing from mid-gestation, two subtracted libraries totaling 2304 cDNA clones were constructed. For achieving this, two reciprocal suppressive/subtractive hybridization procedures (SSH) were applied to early (11 weeks) human placental villi after incubation either in normoxic or in hypoxic conditions. The clones from both libraries (1440 hypoxia-specific and 864 normoxia-specific) were spotted on nylon macroarrays. Complex cDNAs probes prepared from placental villi (either from early pregnancy, after hypoxic or normoxic culture conditions, or near term for controls or pathological placentas) were hybridized to the membranes.
Results
Three hundred and fifty nine clones presenting a hybridization signal above the background were sequenced and shown to correspond to 276 different genes. Nine of these genes are mitochondrial, while 267 are nuclear. Specific expression profiles characteristic of preeclampsia (PE) could be identified, as well as profiles specific of intra-uterine growth retardation (IUGR).
Focusing on the chromosomal distribution of the fraction of genes that responded in at least one hybridization experiment, we could observe a highly significant chromosomal clustering of 54 genes into 8 chromosomal regions, four of which containing imprinted genes. Comparative mapping data indicate that these imprinted clusters are maintained in synteny in mice, and apparently in cattle and pigs, suggesting that the maintenance of such syntenies is requested for achieving a normal placental physiology in eutherian mammals.
Conclusion
We could demonstrate that genes induced in PE were also genes highly expressed under hypoxic conditions (P = 5.10-5), which was not the case for isolated IUGR. Highly expressed placental genes may be in syntenies conserved interspecifically, suggesting that the maintenance of such clusters is requested for achieving a normal placental physiology in eutherian mammals.
==== Body
Background
Preeclampsia (PE) is a severe disorder of human pregnancies affecting up to 10 % of primiparous women in industrialized countries. This hypertensive disease develops from the second half of pregnancy and is associated with proteinuria, and sometimes with oedemas. The causes of preeclampsia are complex and multiple, with a combination of environmental and genetic effects from maternal [1] as well as paternal origin [2,3]. At the histological level, preeclampsia is characterized by a shallow colonization of the maternal endometrium (and more specifically of the wall of the uterine vessels) by the invasive cytotrophoblasts. In a normal gestation, this process occurs during the second trophoblastic colonization wave around the end of the first trimester of gestation allowing the invasive cytotrophoblasts to reach the placental bed. In this case, the arterial wall is infiltrated by endovascular trophoblasts, triggering a suppression of the vasomotor control, thus resulting in a very important dilatation of the lumen and a loss of the elastic properties of the arteries [4-8]. By contrast, when the invasion is defective, remaining too shallow [9], fibrinous material accumulates in the arteries, myocytes proliferate. This may lead to local thrombosis and is therefore supposed to impact on the local oxygen pressure leading to placental ischemia/hypoxia, and ultimately to functional anomalies of the maternal vascular endothelium [10,11]. Finally, the functions of the syncytiotrophoblast, the specific tissue resulting from the fusion of the cytotrophoblasts, may be modified, leading to trophoblast apoptosis. Links between placental pathology and hypoxia are now clearly documented ([12,13], for a recent review, see Challier et Uzan, 2003; [14]). Intra-Uterine Growth Retardation (IUGR) constitutes another group of complex diseases, a large subset of which is associated with placental malfunction, and often with preeclampsia, although this is not at all systematic. IUGR may be defined as a rupture in the normal growth curve of the foetus, although it is generally defined as a birth weight inferior to the 10th or to the 3rd percentile (according to a chosen arbitrary limit) of the smallest babies. This simple definition does not take into account the dynamics of the growth curve, but is very easily workable. This explains at least partly the underlying complexity present under the term IUGR. At least, two very distinct situations are possible: IUGR may be the result of endogenous developmental and growth factors affecting the fetus growth; or, alternatively, a placental defect may inhibit the transfer of nutriments and oxygen from the mother to the fetus (IUGR of vascular origin or not). In the following, vascular IUGR correspond to those exhibiting an abnormal Doppler. In this latter case, the IUGR may be caused by a vascular pathology such as PE. Consequently, there is a clear need for classification of IUGR, and for improving the understanding of its etiology, both issues possibly based upon the characterization of specific marker genes.
In addition, it is now well documented that babies presenting with a small birth weight are at increased risk for developing systemic pathologies at adulthood, such as diabetes or cardiovascular troubles [15,16]. One clear limitation of the therapeutic possibilities for PE and IUGR resides in the fact that whilst many indications suggest that their causes are very precocious during placental development [17], their symptoms occur late, at least from mid-gestation, and more often from the last trimester of the pregnancy. Evaluating the risk early would allow orienting the medical choices towards a better follow-up of the pregnancy, or even towards some pharmacological options, such as the chronic use of mild doses of aspirin.
In order to explore on a wide basis the links between hypoxia and placental diseases, and to identify early putative markers of the preeclamptic pathology, we combined the Subtractive/Suppressive Hybridization (SSH) methodology starting with cDNA material obtained from first trimester placental purified villi with the construction of high-density nylon macroarrays (Figure 1). A first characterization of the libraries has previously been performed by systematic sequencing, enabling to identify all the genes that were modulated by short-term hypoxia in early placental villi, even when this genes are expressed at a low level [18]. However, this systematic sequencing approach does not indicate which gene could constitute an optimal early marker for disease detection (highly and differentially expressed). Therefore, in the present study, we used the less sensitive approach of hybridization in order to identify genes modulated early by hypoxia in pregnancy (the end of first trimester, which is crucial for trophoblast invasion), and also highly expressed and modified in placental diseases occurring later. We report here on the macroarray hybridization results obtained from 20 different sources of cDNA. We demonstrate a clear association between PE and the early hypoxic induction of a series of genes involved in different metabolic processes. This association was totally absent when cDNAs from isolated IUGR were used as probes. We could also demonstrate that the kinetics of hypoxia is not linear, and that genes transcriptionnally modulated in early placentas after 3 h of hypoxia may return to their basal expression level when the culture is maintained for 48 h in hypoxic conditions. Finally, we demonstrate that genes highly expressed in the placenta are clustered in specific chromosome regions, in particular in regions previously defined as containing imprinted genes, such as 11p15.5.
Figure 1 A chart presenting the protocol used to construct the principal tool used in this study: high density nylon membranes spotted with two Suppressive Subtractive Hybridization libraries (SSH). The original cDNAs were obtained from 11-weeks placental villi maintained in normoxia or hypoxia. Two reciprocally subtracted libraries were constructed and spotted at high density on nylon membranes. Then hybridizations were carried out using complex probes from various placentas (either from healthy, or from pathological pregnancies). The rationale of using early villi and hybridizing with near-term villi was the aim of identifying genes modified early by hypoxia, and still modified later chronically in the pathological state.
Results
1. Overview of the genes highly expressed in placental physiology according to their putative interactions
When all the hybridizations were considered, 641 clones out of 2304 yielded a response above the background level of the membrane after an overnight autoradiography. Hybridization with a labeled Oligo dT indicated that 281 clones corresponded to Oligo dT sequences cloned during the SSH procedure. These clones were generally highly labeled in most hybridization experiments, except when the probe was prepared from a placenta where the apparent level of transcriptional activity was dramatically decreased (i.e. vascular IUGR, see below). Finally, 360 sequences corresponded to known genes of which 276 were different (9 mitochondrial and 267 nuclear genes). The normalization effect of the SSH was demonstrated by the fact that 243 genes were found only once. The most frequent was CGA (corresponding to the common alpha chain of four glycoprotein hormones, LH, FSH, TSH and hCG), found in 9 occurrences. The other frequently represented genes corresponded to mitochondrial genes involved in the structure of the mitochondrial ribosome (16S, 6 occurrences, and 12S, 4 occurrences) or in the respiratory chain (COX1, 4 occurrences, COX3, 3 occurrences). The complete set of genes could be divided in 11 distinct cell functions: RNA binding, Protein synthesis, Apoptosis, Inflammation, Cell to Cell contacts, Angiogenesis, Epigenetic mechanisms and Imprinting, Cytoskeleton constitution, Signal transduction, Cell cycle and Lipid metabolism. These functions and the links existing between 117 genes are represented in Figure 2, drawn from literature information. Genes induced by hypoxia are written in red and genes inhibited by hypoxia in blue. Amongst several noticeable features, the picture exhibits a large amount of genes encoding RNA-interacting factors. Several of these genes encode proteins recognizing specific mRNAs. This is the case for NUFIP1 that interacts specifically with EEF1A1 mRNA, this latter encoding a specific elongation factor interacting with ribosomal proteins for elongating nascent polypetidic chain. Similarly, IGFII (Insulin-like Growth Factor II) mRNA Binding Protein-3 (IMP-3), interacts in particular with IGFII mRNA [19], a well known very important actor of placental growth and physiology, as demonstrated recently by the specific invalidation of the IGFII placental isoform [20]. The importance of IMP3 was recently emphasized in a study demonstrating by RNA interference its involvement for enhancing IGFII mRNA translation in K562 leukemia cells [21]. RNA interacting factors represent less than 1% of the total gene content in mammals. In our subset of highly expressed placental genes, they represent over 4% of the total.
Figure 2 Putative physiological relationships between nuclear genes found expressed at a detectable level on the membranes. Amongst a total of 269 nuclear genes, known described relationships could be deduced from the literature for 117 of them. The genes were grouped in 11 categories. In blue and red are represented genes transcriptionnally inhibited or activated by hypoxia, respectively. In green are presented genes that were not detected by hybridization but that may play critical roles in placental physiology. Open boxes present the main physiological action of several of these genes. Arrows indicate an activation effect, while lines terminated by circles indicates an inhibitory effect. Lines terminated by two circles are indicative of a physical interaction between two protein products, or between a protein and a RNA molecule. Table 5 (supplemental table) gives the complete name of the genes displayed on the figure.
2. Hierarchical clustering of genes and hybridization experiments
To take into account the complexity of the tissue, the hybridization experiments were grouped into six categories from the cDNA prepared from samples listed in Table 1. The signals from the different experiments were averaged, as described in Experimental Procedures. The clustering tool developed by Eisen and coworkers [22] was applied to the set of 360 identified genes. Two hundred and seventy five genes could be grouped in seven clusters of similar expression (A to G, figure 3 and 4). The experiments were classified by the program in the following organization: normal placenta, either term or early were grouped into one cluster while placenta from disease states or from villi maintained in hypoxia during 48 h were grouped in another cluster. In a lower order cluster, PE and PER were grouped and associated with early villi maintained during 48 h in hypoxia while isolated IUGR was placed separately. The distribution of genes into two sets (hypoxia-induced versus hypoxia-inhibited) made it possible to assess statistically the possible relationships between the effects on the transcript levels of a relatively short period of hypoxia (3 h) on early placenta (the hypoxic condition that was used to construct the arrays) and pathological states developing later (Table 2). Genes that are specific of the placenta, either early or at term, have a slight trend to be specifically induced by hypoxia in early term placenta. There is also a highly significant tendency of preeclamptic villi, either from isolated PE or PE combined with IUGR, to express genes induced by 3 h hypoxia in early placental villi. The inverted trend is observed at a highly significant level in hybridizations carried out with probes obtained from early term villi maintained in hypoxia during 48 h with an excess of "normoxic" genes found in this situation. We supposed that this observation could be related to a specific kinetics of induction by oxygen concentration, where short exposure to hypoxia may have effects that are at the opposite of long-term hypoxia. Indeed, we could observe that genes whose expression was modified by short periods of hypoxia may later return to their basal expression level as shown by analyzing kinetics of expression in hybridizations with cDNA of the same placenta maintained in hypoxia during 3, 24 or 48 hours (Figure 5). In the cases of isolated vascular IUGR, due to a drastic limitation of the materno-foetal blood flow only 40 positive clones yielded a detectable signal (instead of 400–500 when other probes were hybridized to the membranes, including the polyA containing clones). These positive clones corresponded to mitochondrial genes, IGFII and PSG4, PSG5, and PSG7 (Pregnancy Specific Glycoproteins 4, 5 and 7), indicating that these genes constitute a minimal survival set to sustain gestation. Moreover, the signal level of positive clones was quite comparable with that of other experiments, demonstrating that the relevant RNA species are indeed present at a high level. Since transcriptional activity is very reduced in vascular IUGR (as shown by the very low abundance of polyA+ RNAs), this suggests that some mRNA molecules are specifically protected from degradation in the very harsh pathological condition of vascular IUGR.
Table 1 Status of the patients used in the study
Patient Status RT (*) Weeks Amenhorrhea Year mother HTA (**) Proteinuria g/24 h Uterine Doppler (****) Oligoamnios Apgar 1 Apgar 5 SEX Birthweight in g (Percentile)
3008 Term (control) 784 38+4 32 normal normal normal normal 10 10 M 3050
7497 Term (control) 816 38+0 34 normal normal normal normal 10 10 M 3270
497 Term (control) 817 40+5 35 normal normal normal normal 10 10 M 3040
3007 Term (control) 818 38+4 43 normal normal normal normal 10 10 M 3430
3011 Term (control) 819 38+0 37 normal normal normal normal 10 10 M 3710
3013 Term (control) 820 39+4 42 normal normal normal normal 10 10 F 2740
3017 Term (control) 821 38+5 41 normal normal normal normal 10 10 F 3110
3004 PE 765 28+5 34 205/120 10,69 normal normal 2 8 M 830
3005 PE 766 37+4 45 140/90 0,17 Unilateral normal 10 10 F 2930
225 PE 807 27+0 24 145/90 2,7 normal normal 9 10 M 1080
242 PE 867 33+1 26 150/95 5,36 normal normal 8 10 M 1960
3010 PE 808 32+5 36 160/110 14,36 normal normal 10 10 M 1196
50 PE 885 30+0 22 175/105 8,68 normal normal 8 10 M 1390
4003 PE+IUGR 915 34+1 36 180/120 2,1 normal amniotic fluid in excess 0 7 F 1670 (< 5)
3012 PE+IUGR 770 34+1 37 140/90 6,98 normal unknown 9 10 F 1800 (between 5 and 10)
3016 PE+IUGR 772 28+0 37 160/100 nc bilateral Abnormal 7 10 M 780 (< 5)
3003 vascular IUGR 773 37+3 26 160/90 normal bilateral unknown 10 10 M 1890 (between 5 and 10)
3021 vascular IUGR 810 31+3 35 normal normal bilateral Abnormal 7 10 M 1380 (> 10 but ***)
3022 vascular IUGR 814 37+0 33 normal normal unknown Abnormal 3 9 M 2330 (between 5 and 10)
* RT: reversed transcribed RNA sample
** HTA: Arterial Hypertension
*** IUGR defined bya break in the intrauterine growth curve, a bilateral doppler associated with an absence of diastolic pressure
**** Presence of a doppler signal in one or both the umbilical arteries
Figure 3 Pictures obtained after data clustering of the SSH hybridizations using the Treeview software [22]. The programs were used according to the parameters described in Material and Methods. Clusters of genes expressed in specific situations are represented. Above the general tree are presented the means of the different hybidization grouped into 6 categories of probes used. Clusters of transcriptionnally induced genes could be characterized. A, Full Term Placentas (Mean FTP); B, Early Term Placentas (Mean ETP), C, Preeclampsia with IUGR (mean PER); D, PER + isolated PE; E, isolated IUGR (Mean R), F, isolated PE (Mean PE), G, 48 h hypoxia.
Figure 4 Pictures obtained after data clustering of the SSH hybridizations using the Treeview software [22]. The programs were used according to the parameters described in Material and Methods. Clusters of genes expressed in specific situations are represented. Above the general tree are presented the means of the different hybidization grouped into 6 categories of probes used. Clusters of transcriptionnally induced genes could be characterized. A, Full Term Placentas (Mean FTP); B, Early Term Placentas (Mean ETP), C, Preeclampsia with IUGR (mean PER); D, PER + isolated PE; E, isolated IUGR (Mean R), F, isolated PE (Mean PE), G, 48 h hypoxia.
Figure 5 Kinetics of hypoxia regulation in early (11 weeks) placentas. Six examples of genes exhibiting a transcriptional arrest under short hypoxic conditions, but coming back to almost normal levels of expression under extended hypoxic conditions.
Table 2 Statistics of gene induction under various conditions
Category Normoxia (plates 1–9) Hypoxia (plates 10–24) Chi2 Observations
A Full-Term Placenta Observed 12 26
Expected 14,25 23,75 0,451 38
B Early+Full-Term Placenta Observed 9 31
Expected 15 25 0,050 40
C PE+RCIU Observed 9 37
Expected 17,25 28,75 0,012 46
D PE+RCIU and isolated PE Observed 10 10
Expected 7,5 12,5 0,248 20
E RCIU Observed 6 9
Expected 5,625 9,375 0,841 15
F Isolated PE Observed 2 35
Expected 13,875 23,125 0,000 37
G Early Term Placenta Hypoxia 48 h Observed 45 36
Expected 30,375 50,625 0,001 81
Each category (A-G) corresponds to a cluster of genes observed in figure 3 (a-b). The expected values were calculated from the proportion of clones in each subtracted library. Significant Chi2 values are represented in bold characters
3. Identification of factors specific of the pathological status of the placenta
Table 3 was extracted from the database using PE as a keyword in the field "Maximal signal". It contains 56 different genes, amongst a total of 71. Among these genes, some were observed in only one PE case (27), while some others were observed in several or all the hybridizations with probes obtained from PE (29 signals), 6 in PE +IUGR, 6 only in the case of a severe PE that was used in the hybridization (clinically defined as an arterial hypertension exceeding 160 mm Hg), while the 3 left could also be observed at a high level in normal term placentas. The set of genes highly expressed in more than one PE constitutes of course a collection of natural candidates for a further exploration of the pathology. These 17 genes are: ACTG1, ATP5B, ATP6, ANGPTL4, CGA, COX1, COX3, CSHL1 (human Placental lactogen), GAPDH, FLJ22728, H19, ND1, ND3, NUFIP1, PSG5, PSG7 and RPL41. Some of these genes, such as CSHL1 have already been identified as PE markers [23,24]. Some others such as ANGPTL4, COX1, FLJ22728, H19 and NUFIP1 are completely new candidates. There were 22 signals corresponding to genes highly elevated in IUGR corresponding to 20 different factors, 15 of which were not correlated with PE (Table 4).
Table 3 Genes induced in PE
Gene symbol Gene name Maximal signal Protein category Chromosomal localization Library address
16S ribosomal RNA 16S ribosomal RNA one PE case Transcription/Translation Mitochondrie 18B8
16S ribosomal RNA 16S ribosomal RNA PE+IUGR Mitochondrial metabolism Mitochondrie 14H2
18S rRNA ARN 18S severe PE Transcription/Translation 13F12
ACTG1 actin, gamma 1 several PE Structure protein 17q25 22G1
angiopoietin-like 4 ANGPTL4 = PPARG angiopoietin related protein several PE Transcription/Translation 19p13.3 11D9
ATP5B ATP synthase, H+ transporting, mitochondrial F1 complex, beta polypeptide several PE Mitochondrial metabolism 12p13-qter 11A4
ATP6 ATP synthase F0 subunit 6 several PE Mitochondrial metabolism Mitochondrie 16D7
C9orf90 chromosome 9 open reading frame 90 DKFZp762G199 one PE case 9q34.13 2C10
CDC37 CDC37 cell division cycle 37 homolog (S. cerevisiae) several PE Cell cycle 19p13.2 10E10
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 23B10
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 2C5
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 23D12
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 10H5
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 2H4
CGA glycoprotein hormones, alpha polypeptide several PE Signal transduction 6q12-21 16D5
CGA glycoprotein hormones, alpha polypeptide Terms/PE Signal transduction 6q12-21 8B8
COX1 Cytochrome c oxidase subunit I several PE Mitochondrial metabolism Mitochondrie 13H7
COX1 Cytochrome c oxidase subunit I several PE Mitochondrial metabolism Mitochondrie 15B3
COX1 Cytochrome c oxidase subunit I several PE Mitochondrial metabolism Mitochondrie 12H9
COX1 Cytochrome c oxidase subunit I several PE Mitochondrial metabolism Mitochondrie 18D6
COX2 Cytochrome c oxidase subunit II one PE case Mitochondrial metabolism Mitochondrie 8B7
COX3 Cytochrome c oxidase subunit III several PE Mitochondrial metabolism Mitochondrie 20F4
CSHL1 chorionic somatomammotropin hormone-like 1 several PE Signal transduction 17q24.2 19A7
CSNK1A1 Casein kinase 1, alpha 1 one PE case Signal transduction 5q32 18A8
DDX3X DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, X-linked one PE case Transcription/Translation/Modifi p11.3-11.23 17F11
DKFZP434F2021 DKFZP434F2021 protein one PE case 3q13.2 6A3
DKFZp586F1223 Hs.28540 severe PE 11q23 20F7
EEF1A1 eukaryotic translation elongation factor 1 alpha 1 one PE case Transcription/Translation 6q14.1 7A11
EIF4B eukaryotic translation initiation factor 4B one PE case Transcription/Translation 12q13.13 2C7
ERVWE1 endogenous retroviral family W, env(C7), member 1 (syncytin) one PE case Cell-cell contacts 7q21-22 8G1
FEM1C fem-1 homolog c (C.elegans) PE+IUGR Transcription/Translation 5q22 16D6
FEM1C fem-1 homolog c (C.elegans) PE+IUGR Transcription/Translation 5q22 8D9
FLJ11149 riboflavin kinase one PE case Transport 9q21.31 23F12
FLJ22728 hypothetical protein FLJ22728 DKFZp761I1913 several PE Transport 11p15.2 23G6
FLJ22728 hypothetical protein FLJ22728 DKFZp761I1913 several PE Transport 11p15.2 23B6
GAPD glyceraldehyde-3-phosphate dehydrogenase several PE Mitochondrial metabolism 12p13 23D3
GAPD glyceraldehyde-3-phosphate dehydrogenase several PE Mitochondrial metabolism 12p13 24B3
GLIPR1 GLI pathogenesis-related 1 (glioma) PE+IUGR 12q21.1 16G10
H19 H19, imprinted maternally expressed untranslated mRNA several PE RNA gene 11p15.5 18C1
H3F3B H3 histone, family 3B (H3.3B) severe PE Chromatin structure 17q25 11C8
IL8 interleukin 8 severe PE Apoptose regulation 4q13-q21 12G11
ITGAV integrin, alpha V (vitronectin receptor, alpha polypeptide, antigen CD51) one PE case Cell-cell contacts 2q31-32 11B11
LAMA5 laminin, alpha 5 one PE case Apoptose regulation 20q13.2-13.3 1C3
LOC126731 LOC126731 Terms/PE 1q42.13 24F2
LOC374933 Homo sapiens LOC374933 (LOC374933), PE+IUGR 1p36.33 21G10
MAN1 integral inner nuclear membrane protein one PE case Transcription/Translation/Modifi 12q14 7A7
MATR3 Matrin 3 severe PE Transcription/Translation/Modifi 5q31.3 3C8
MGC2714 hypothetical protein MGC2714 PE+IUGR 11q22.2 13C7
ND1 NADH dehydrogenase subunit 1 several PE Mitochondrial metabolism Mitochondrie 12C4
ND3 NADH dehydrogenase subunit 3 several PE Mitochondrial metabolism Mitochondrie 18H7
NRCAM neuronal cell adhesion molecule one PE case Cell-cell contacts 7q31.1-q31.2 18F11
NUFIP1 nuclear fragile × mental retardation protein interacting protein 1 several PE RNA-interacting factor 13q14 4E6
OSBPL2 oxysterol binding protein-like 2 one PE case Signal transduction 20q13.3 1C3
PSG4 pregnancy specific beta-1-glycoprotein 4 one PE case Signal transduction 19q13.2 16C9
PSG4 pregnancy specific beta-1-glycoprotein 4 one PE case Signal transduction 19q13.2 11A9
PSG5 pregnancy specific beta-1-glycoprotein 5 several PE Signal transduction 19q13.2 10F12
PSG7 pregnancy specific beta-1-glycoprotein 7 several PE Signal transduction 19q13.2 10C11
PSG7 pregnancy specific beta-1-glycoprotein 7 several PE Signal transduction 19q13.2 16F9
RPL41 ribosomal protein L41 several PE Transcription/Translation/Modifi 12q13 18B9
RPS11 ribosomal protein S11 one PE case Transcription/Translation/Modifi 19q13.3 13C9
RPS24 ribosomal protein S24 one PE case Transcription/Translation/Modifi 10q22-23 4C8
RPS6KC1 ribosomal protein S6 kinase, 52kDa, polypeptide 1 severe PE Transcription/Translation/Modifi 1q41 10C8
rRNA 28S Human 28S ribosomal RNA gene one PE case RNA gene 8q21.1-q21.2 6G8
S100P S100 calcium binding protein P one PE case Cell cycle 4p16 5D3
SMARCC1 SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily c, member 1 one PE case Structure de la chromatine 3p23-21 8A2
SRP9 signal recognition particle 9kDa one PE case Transport 1q42.13 5E10
UCP2 uncoupling protein 2 (mitochondrial, proton carrier) one PE case Mitochondrial metabolism 11q13 8C3
VIM vimentin one PE case Structural protein 10p13 5H2
WBSCR1 Williams-Beuren syndrome chromosome region 1 EIF4H one PE case Transcription/Translation/Modifi 7q11.23 14E4
WBSCR16 Williams-Beuren syndrome chromosome region 16 one PE case Transcription/Translation/Modifi 7q11.23 12C8
YWHAZ tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta polypeptide Terms/PE Signal transduction 8q23.1 4B6
Table 4 Genes induced in IUGR
Gene symbol Gene name Maximal signal Protein category Chromosomal localization Library address
16S ribosomal RNA 16S ribosomal RNA PE+IUGR Mitochondrial metabolism Mitochondria 14H2
BC014242 Hs.5064 IUGR 5 10C6
cDNA DKFZp686G03142 Homo sapiens mRNA; cDNA DKFZp686G03142 IUGR 5 16G12
COTL1 coactosin-like 1 (Dictyostelium) IUGR Structural protein 16q24.1 11A7
CUL4B cullin 4B IUGR Cell cycle Xq23 2D5
DAF decay accelerating factor for complement (CD55, Cromer blood group system) IUGR Signal transduction 1q32 13G10
FEM1C fem-1 homolog c (C.elegans) IUGR+PE Transcription/Translation 5q22 8D9
FEM1C fem-1 homolog c (C.elegans) IUGR+PE Transcription/Translation 5q22 16D6
FTH1 ferritin, heavy polypeptide 1 IUGR Transport 11q13 20D12
FTH1 ferritin, heavy polypeptide 1 IUGR Transport 11q13 2G1
GLIPR1 GLI pathogenesis-related 1 (glioma) IUGR+PE 12q21.1 16G10
IMAGE:3453987 Homo sapiens cDNA clone IMAGE:3453987 IUGR 4 10C5
IGF2 Insulin-like growth factor 2 (somatomedin A) IUGR Growth factor 11p15.5 15E3
IMP-3 IGF-II mRNA-binding protein 3 IUGR Transcription/Translation/Modification 7p11 18G10
KIAA1354 KIAA1354 protein IUGR Signal transduction 9p22 2D4
LOC285429 hypothetical protein LOC285429 IUGR 4p14 3B7
LOC374933 Homo sapiens LOC374933 (LOC374933), PE+IUGR 1p36.33 21G10
MGC2714 hypothetical protein MGC2714 PE+IUGR 11q22.2 13C7
ND1 NADH dehydrogenase subunit 2 IUGR Mitochondrial metabolism Mitochondria 15H1
PAPOLA poly(A) polymerase alpha IUGR Transcription/Translation/Modification 14q32.31 10A7
SND1 staphylococcal nuclease domain containing 1 EBNA2 coactivator p100 IUGR Transcription/Translation 7q31.3 2D4
TPI1 triosephosphate isomerase 1 IUGR Transport 12p13 8D2
4. Promoter structure of PE/Hypoxia induced genes
Sequences spanning from 5000 bp of DNA 5' of the transcription start site and extending 200 bp after, were recovered from GenBank at the NCBI and were analysed for their composition in CpG islands and the presence or absence of Hypoxia Inducible Factor 1 α (HIF1α) binding sites. HIF is a transcription factor protected from degradation in many cell system under hypoxic conditions. It plays an essential role in modulating responses to hypoxia by inducing or inhibiting multiple genes. Four possible binding sites have been described for this factor, ACGTGC, ACGTGG, GCGTGC and GCGTGG [25].
The promoters could be classified into four categories: Several of them, such as the promoters of PSG4, 5 or 7, CUL4, or CGA, do not contain any noticeable CpG islands. Several others contain a concentration of CpG islands very close to, or encompassing the transcription start site (PAPOLA, WBSCR1, COTL1, SMARCC1). Other promoters contain CpG islands at around 2000 bp 5' upstream of the ATG (FEM1C, FLJ11149). Finally, some promoters are highly enriched in CpG islands over the whole 5000 bp examined (ACTG1, IGFII, H19). The set of promoters with the CpG density and the position of putative HIF binding sites are supplied as Supplemental data file 1. We could also classify these promoters according to the maximal density of CpG achieved in the 5 kb window analysed in five groups (less than 50, from 50 to 100, 100 to 150, 150 to 200 and more than 200). Consistently with the high GC-richness of putative HIF binding site, there was a clear linear relationship between the CpG density and the number of putative HIF binding sites (R2 = 0.86). The distribution of these sites was analysed inside the promoters by sharing these promoters into two halves of equal length and counting the putative sites inside each subsequence. A student T-test did not reveal any preference towards one side against the other (P = 0.45).
5. Non-random chromosome distribution of placental genes in eutherian mammals
The analysis of the cytogenetic location available from our database revealed that several genes expressed at a detectable level in hybridization experiments were clustered to specific chromosomal regions. We focused our interest on the precise chromosome location of these genes, using the information available at the NCBI site [26] in order to obtain their precise physical position expressed in megabases.
We could confirm by statistical analysis (see Experimental Procedures) that the distribution of the subset of highly expressed placental genes was not random. On the contrary, we identified 8 clusters on 7 different chromosome regions: 1p36 (9.76 Mb), 6q14 (0.23 Mb), 11p15 (11.63 Mb), 11q13 (16.28 Mb), 12q13 (3.65 Mb), 19q13 (12.35 Mb), 20q13 (15.11 Mb) and Xq24 (7.45 Mb) (Figure 6). Among these regions, 4 are known to contain imprinted genes (1p36, 11p15, 19q13 and 20q13). These four regions are conserved in synteny and colinearity in mice, while the regions corresponding to human 11q13 and 12q13 are separated on different mice chromosomes. Despite the more limited mapping information available in pigs, obtained at the Iccare website [27] and cattle [28,29], we did not detect any chromosome breakpoints in these species for these specific chromosome regions.
Figure 6 Non-random chromosomal clustering of genes highly expressed in the placenta. Statistical analysis of intergenic distances revealed the existence of 8 clusters distributed on 7 chromosomes. The clusters located on 1q36, 11p15.5, 19q13 and 20q13 are known to belong to imprinted chromosomal regions.
Discussion
In this study we designed a new transcriptome resource directed at evaluating the effects of hypoxia on human placenta. This tool is particularly original if compared to the existing commercial membranes, since the distribution of clones into two subgroups makes it possible to analyze statistically whether one condition is connected to hypoxia-induced stimulation or inhibition of gene expression. This was clearly shown for PE, a pathologic condition that correlated very well with the induction of "hypoxic" genes. Since the placenta is one of the organs presenting the greatest abundance of diversified transcripts, these membranes can also be useful for characterizing either other biological systems, or the effects of hypoxia on other tissues. The use of the SSH approach to construct the membranes may also generate some biases as the cloning is dependent on the existence of Rsa1 restriction sites in the starting material (driver and tester cDNAs). Nevertheless, our tool, focused on early hypoxia, is a useful complement to other DNA arrays experiments, based on commercial membranes [30]. As shown in our study, this tool was used to analyze at the mRNA level the consequences of the two most frequent placental pathologies, preeclampsia and Intra-Uterine Growth Retardation.
Some of the genes found associated with PE in the present study had already been evaluated as putative markers of placental pathological status, such as human chorionic gonadotrophin (hCG) and human placental lactogen (hPL, also known as CSHL1), as well as pregnancy specific glycoproteins (PSGs). A significant serum increase in hCG was found more prevalent in preeclamptic women [23]. In another recent study, hPL and PSGs were found reduced at 17 weeks in the serum of patients who develop later a preeclamptic condition, albeit serum levels are restored later [31]. hCG is composed of two polypeptidic chains, α and β, encoded by CGA (common to four polypeptide hormones) and CGB, respectively. It is known to play important roles in placental physiology [32]. In our study CGA was indeed found induced in several PE cases, but could not be associated with a specific differential oxygen status. CGB was present in the membranes, but was not specifically induced in preeclampsia. The only CGB clone was located in plate 7H12 (normoxia). The over-expression of CGA could lead to an overall increase of hCG in the patient serum however the mechanisms involved for regulating the expression of the two polypeptidic chains constitutive of hCG seem to display opposite modes of regulation.
Recently, Bersinger and Odegard [31] have demonstrated that in IUGR, hPL is continuously lowered later in pregnancy. This is consistent with our results, as we could not detect any trace of expression of this gene in the IUGR probes that were hybridized on the membranes. Similarly, the expression level was low in control term placentas (FTP). However, hPL appeared strongly induced in PE, especially in severe cases. These results at the mRNA level differ from the findings of Bersinger and Odegard [31] concerning the hPL protein levels, which were going back to normal at weeks 28 and 33 of preeclampsia. Possibly, the accumulation of hPL mRNA would not be followed by translation in PE.
Only three PSG were found in the membranes amongst the 11 existing genes. PSGs are supposed to play an essential immunomodulatory effect in pregnancy [33-35]. Consistently with our result, in a recent microarray study [36] the authors used a commercial membrane to identify differentially expressed genes in preeclamptic and normal placentas. Among the ten most highly expressed genes in the membranes, the authors have identified PSG4, PSG5 and PSG7. Interestingly, those are exactly the highly expressed genes that we have found in term placentas suggesting that only these PSGs are specifically expressed at a high level in term pregnancies. This observation raises the question of the regulation of the entire cluster of PSG genes which spans roughly 550 kilobases on 19q13 in the order 3, 1, 6, 7, 11, 2, 5, 4, 9 (the precise localization of the two other genes PSG 8 and PSG10 is not yet known). To address the question of PSG regulation, we performed a Clustal alignment of 3 kilobases upstream of the first codons of PSG 1, 2, 3, 4, 5, 6, 7, 9, and 10. The clustering of the promoters indicated a very high level of conservation and was not able to group PSG4, PSG5 and PSG7 as more similar together (not shown). This indicates that the specifically high expression level of 3 genes, which are not contiguous, depends on specific long-range acting chromatin factors.
ANGPTL4, also called FIAF (fasting-induced adipose factor) is another 3 h hypoxia-activated gene recurrently found in PE cases. It is a downstream target of PPARγ (peroxisome-proliferator-activated receptor γ), and is therefore supposed to regulate lipid metabolism and glucose homeostasis [37]. Although belonging to a family of 9 genes, it is the only one that we could detect by hybridization, suggesting a highly specific mode of regulation, consistently with what has been described of its specific regulation compared to ANGPTL3 [38]. In mice, ANGPTL4/FIAF is increased in the plasma by fasting and decreased by high fat diet [39], demonstrating its involvement in lipid capture in difficult physiological conditions, of which PE may represent a paradigmatic case.
Another unexpected actor of the onset of PE could be H19. This RNA-encoding gene is of ill-defined function. It is located in the 11p15.5 cluster of imprinted genes and expressed by the maternal allele, in apparent opposition with IGFII. Both genes appear regulated by a common DMR (differentially methylated region), and have opposite effects on fetal growth in mice [20,40]. Both IGFII and H19 were found in the set of hypoxia-induced genes, consistently with the existence of HIF1α binding sites in their promoters suggesting the ability to respond almost instantaneously to variations in oxygen concentration. For IGFII, this is also consistent with its described although controversed angiogenic properties [41,42]. However, while IGFII was not PE-specific (and was found expressed in term placentas, pathologic placentas (either PE or vascular IUGR) as well as in early placentas exposed to hypoxia), H19 was specifically expressed in PE and strongly in severe PE.
In our study, we could observe that true vascular IUGR results in a drastic reduction of the transcriptional activity of the placenta. In these cases, only mRNAs for IGFII, PSG4, PSG5 and PSG7, and mitochondrial genes were present, and no polyA could be detected. Therefore, we could conclude that in vascular IUGR, only a minimal set of genes was transcriptionnally maintained in the placental tissue in order to prevent spontaneous abortion, (i) transcripts from genes of the respiratory cascade, (ii) IGFII, one isoform of which is the essential growth factor in placenta [20], and (iii) three genes of the PSG cluster. The absence of polyA in the cDNAs from purely vascular IUGR suggests that the remaining clones are maintained by stabilization of the transcripts rather than at the transcriptional level. This may be substantiated by the observation of IMP3cDNA in the transcripts specifically expressed in less severe cases of IUGR. IMP3 encodes a factor interacting with the IGFII mRNA, possibly stabilizing the transcript, and belonging to an imprinted region on chromosome 7.
In non-purely vascular IUGR, FTH1 (ferritin) was expressed at a high level, suggesting the existence of genetic adaptative mechanism to a restricted supply of nutrients. However, ferritin was found twice in the library, without correlation with the oxygen status, showing that this gene does not participate directly in the rapid response to low oxygen concentrations. Ferritin is a major factor for ensuring a sufficient iron store to the neonate at birth. This echoes to a study showing that children presenting a low iron store at birth had low serum ferritin concentrations at 9 months [43], suggesting a risk of iron deficiency in the second postnatal year.
Interestingly, in both pathologies, we found the induction of FEM1-C. This gene, discovered as a homologous of a Caeborhabditis elegans gene contains KH domains, themselves highly present inside FGIF, the principal inducer of the foetal globin. Again, its presence in the library was not correlated with short hypoxia (3 h). However, we observed that its mRNA concentration was lowered at 48 h hypoxia. Its occurrence in pathological situations could refer to a specific adaptative mechanism aiming at increasing the oxygen capture for the foetus. Its induction fits well with the observed increase in ferritin mRNA, both genes FEM1-C and FTH1, aiming at building the proteic and prosthetic part of the globin polypeptide, respectively.
Some particularly appealing genes for the diseases studied were present in the membranes but were not revealed by hybridization with specific pathologic samples, as they appear down-regulated in pathological conditions. Among those, we found two clones corresponding to SERPINE2 and one to SERPING1. These two genes are inhibitors of Serine-Proteases that may play a role in thrombus clearance, and therefore are necessary for adequate circulatory functions. SERPING1 (also known as C1-inhibitor) is particularly interesting, since several mutations of this gene are involved in the development of hereditary angioedema (MIM&606860, [44-47]). Oedemas frequently accompanies the preeclamptic condition, therefore specific malfunctions of SERPINE2 and SERPING1 could represent risk factors for placental diseases. SERPINE2 also appeared in our previous study [18] as a gene present in the hypoxic library at the highest number of occurrences. Therefore, it suggests that a gene induced by short-term hypoxia, may be on the contrary down-regulated by chronic hypoxia, such as observed in PE. Recently, we have shown that the same type of regulation is also true for SERPING1 (F. Quetin and S. T Chelbi, unpublished results).
We performed a systematic analysis of the 5' regions of the PE and IUGR induced genes. In many of them, we found CpG islands that may be modulators of gene expression. In a future work we shall analyze several of these regions by bisulphite analysis of CpG methylation in normal and pathological cases, to try to give a molecular basis to the observed differences in gene expression. In most cases, we discovered more than one HIF binding site in the various promoters identified. Experimental verification will be needed to evaluate the significance of these binding sites.
A surprising result in our study was the demonstration that genes expressed at a level sufficiently high to be detectable in one at least of our hybridization experiments, are mapped to specific chromosome regions. This is particularly the case for clusters of imprinted genes. In these respects, the involvement of the 11p15.5 cluster in the physiopathology of PE fits well with the recent observation that the invalidation of p57/kip2, another gene of this cluster results in "preeclampsia-like" symptoms in mice [48].
In recent studies, several groups have demonstrated that the human genome is organized in large clusters of highly expressed genes [49,50]. This high level organization of the human genome is conserved in other mammalian species, such as cattle and pigs [51]. In the present study, we demonstrate that beside expression level, clustering also exists for functional purposes, such as placental physiology. This vision of the mammalian genome is consistent with the hypotheses and experimental demonstrations developed by Cremer and co-workers, indicating that the presence of entire chromosomes or chromosome regions inside sub-compartments of the nucleoplasm triggers variations in expression levels [52].
Conclusion
In conclusion, our study has:
• Demonstrated highly significant differences between isolated IUGR and PE concerning the effects of placental oxygen pressure on gene expression. This finding suggests that very different mechanisms are involved when IUGR originates from a fetus-borne developmental dysfunction, and when IUGR results from a vascular defect such as PE.
• Provided the scientific community with a directly available tool, making the link between hypoxia and placental diseases
• Confirmed the importance of several risk factors for PE and IUGR (such as CGA, PSGs and hPL)
• Suggested new possible targets for diagnosing early these pathologies, and possibly for alleviating their effects (such as COX1, ANGPTL4, H19, FTH1, FEM1c, IMP3, SERPING1 and SERPINE2). Since the SSH was carried out from an early placenta, our study makes the link between potential markers of early oxygen depletion, and the late development of preeclamptic lesions, strengthening the idea that PE may be caused by early alterations of placental function.
• Demonstrated the existence of a genomic organization of placental function in placental mammals.
This work opens the way to characterize at the molecular level the physio-pathological mechanisms of very complex situations that often perturb normal pregnancies.
Methods
Patients and ethics
All the placentas from the patients were collected from four Parisian maternities (Cochin, St Antoine, Institut de Puériculture and St Vincent de Paul). This study was approved by the Ethics Committee of Paris Cochin (France), CCPPRB (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale). All the patients have given their written consent for the use of their placenta. Early term placentas (ETP) were obtained from healthy women undergoing legal abortion by vacuum curettage between 8 and 12 weeks of amenorrhea. "Late" placentas were obtained from caesarean section outside labor from healthy mothers ("FTP or Full-Term Placenta", between 38 and 39 weeks of amenorrhea) or from mothers with pathological pregnancies ("PE" or "IUGR", between 28 and 37 weeks of amenorrhea). The details about the patients used are given on Table 1.
Placental villi from term placentas (control and pathological)
Biopsy samples were rapidly collected at six to ten various locations from each placenta between the decidual and chorionic plates. Villous tissue was dissected free of fetal membranes, vessels and tissue from maternal origin, rinsed and minced in Ca2+, Mg2+ free Hank's Balanced Salt Solution (HBSS).
Placental villi from early placentas and hypoxia conditions
Floating villi isolated by fine mechanical dissection were cleaned in order to remove fetal membranes, large vessels and tissue from maternal origin. They were rinsed and minced in Ca2+, Mg2+ free Hank's Balanced Salt Solution (HBSS). They were plated on 60 mm diameter dishes (0.4 g villi/dish) in 3 ml of RPMI 1640 medium supplemented with 10% (V/V) fetal bovine serum, 25 mM HEPES, 2 mM glutamine and antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin). They were made hypoxic by placing them in a Lwoff chamber at 37°C and exposed to an oxygen-depleted atmosphere (2% O2, 5% CO2, 93% N2) or maintained at 37°C in normal conditions in humidified 5%CO2-95% air during 3 or 48 h. The hypoxia was controlled by checking the atmospheric oxygen pressure at the end of the experiment using an ABL725 gas analyzer (Radiometer, Copenhagen) as previously described extensively [18]).
After incubation or alternatively, directly after dissection, villi from either early or term placentas were dry-frozen in Trizol™ reagent (Life Technologies, Cergy, France) for RNA isolation, and store at -80°C until processed.
RNA isolation, Poly-A+ preparation, cDNA synthesis and Subtraction experiment
For the SSH experiment [53], the starting material was constituted by villi from a 11 week normal placenta. After dissection, 2 batches of 2 g of villi from different zones of the placenta were cultured for 3 h either in hypoxic or normoxic conditions. Total RNA was extracted from villous tissue using Trizol™ reagent according to the method of Chomczynski and Sacchi [54]. PolyA+ RNAs were then fractionated from total RNA on Oligo dT latex beads using Macherey-Nagel columns (Macherey-Nagel, Germany). cDNA were synthesized from 1.5 – 2 μg of polyA+ using the reverse transcriptase of the cDNA-select kit (PCR-Select cDNA Subtraction Kit, CLONTECH). Linkers were ligated to one batch of each sample ("tester") and hybridized with an excess of the other sample ("driver") using two consecutive hybridization steps. The products obtained are then amplified by PCR using two nested primers present in the linkers. After the SSH procedure, two effects are expected to happen on the cDNAs: firstly, a normalization that reduces drastically the number of molecules corresponding to highly abundant mRNA species, and secondly, a subtraction which enriches considerably each "tester" in tester-specific molecules, i.e. molecules that were initially rare in the "driver". The experiment was then performed by following accurately the manufacturer's advices, and performing all the possible controls at each step. The normalization was monitored by measuring the GAPDH level by quantitative RT-PCR in the SSH product versus a mock-subtracted sample. In our case, we could show that the normoxia specific product (called "N-H") was roughly depleted 4,000 times in GAPDH, while the hypoxia-specific product (called "H-N") was depleted around 3,000 times in GAPDH. The quality of the subtraction was evaluated by RT-PCR using primers for the Vascular Endothelial Growth Factor (VEGF) (sense: 5'-ATGAACTTTCTGCTGTCTTGGGTG-3' and antisense: 5'-CTCACCGCCTCGGCTTGTCAC-3'), this gene being highly induced in hypoxic condition. While a band was clearly detectable after 25 PCR cycles in the hypoxia versus normoxia SSH product, no signal could be seen after 40 PCR cycles in the normoxia versus hypoxia SSH product (not shown).
Library construction and spotting on high-density macroarrays
The secondary PCR product (nested PCR at the end of each subtraction) was cloned in pGEMT vector (Promega corp.), by incubating 300 ng of each of the PCR products with 50 ng of vector with T4 DNA ligase (Biolabs) and the appropriate buffer during 48 h at 4°C, in a total volume of 20 μl. The ligations were diluted to the fifth in sterile water and used to transform DH10B electrocompetent E. coli bacteria using a Biolabs electroporator. The transformation titre was evaluated on 8 cm LB-agar plates prepared with 100 μg/ml ampicillin and IPTG/XGal [55], and the next day ~1500 colonies from each of the two subtractions were plated on 22 cm × 22 cm LB-agar trays. After an overnight growth, the colonies were manually picked (864 from the N-H library and 1440 from the H-N library) and grown individually in 96-well mega plates containing 1 ml LB/100 μg/ml ampicillin and 10% glycerol. The plates were covered with porous covers and grown overnight at 37°C under rocking at 200 rpm. Part of the culture was stored at -80°C and 200 μl were transferred to Genetix™ 96-well plates and used by a robotic station for double-spotting on nylon membranes (Hybond N+, Amersham) overlaying new LB-agar/ampicillin trays. The cultures were grown overnight on the membranes at 37°C and the DNA was prepared in situ by three consecutive incubations in NaOH 0.5 M (20 min.), Tris-Cl 1.5 M (10 min.), pH 7.5 and SSC 2X (10 min.). Then, the DNA was fixed to the membranes by UV-light irradiation during 5 min (304 nm).
Probes, Probe labelling and Hybridizations
Early placentas around 11 weeks (2 hybridizations), early placentas exposed to an hypoxic environment during 48 hours (2 hybridizations), or term placentas from normal (3 hybridizations, out of which one correspond to a mixture of 6 term cDNAs) and pathological states placenta (10 PE, 2 PE+IUGR, 3 IUGR, one of which was a mixture of 2 IUGR were hybridized to two sets of membranes. As the clones were spotted in duplicates, each hybridization resulted in four positive signals for an expressed gene. Probes were synthesized from 4 μg of total RNA by random hexanucleotide priming, using 20 μM of primers in the presence of Moloney Murine Leukemia Virus Reverse Transcriptase (MMLV) in a total volume of 25 μl at 39°C according to the manufacturer's specifications (Life Technologies). The cDNA products were stored at -20°C until required for probe labelling. Labelling was performed with 50 ng of cDNA and 5 μl α33P dATP using the Biolabs Klenow labelling kit, and allowing the reaction to last during four hours, following the manufacturer's recommendations for 33P labelled nucleotides. The ratio of radiolabeled nucleotide incorporation was evaluated by counting before and after TCA precipitation, and was systematically close to 80% [55]. Hybridizations were performed overnight with a standardized amount of radiolabelled probe, in a hybridization oven at 42°C in the following buffer: SSC 4X, Denhardt's 2.5X, SDS, 0.5%, Tris HCL-EDTA (10 mM-1 mM, pH 7.5), Dextran sulfate 1 g/10 ml. The membranes were then washed three times at 58°C in 2X SSC, wrapped in Saran™ and autoradiographed overnight.
Sequencing
Addresses of positive clones were identified from the autoradiography and the corresponding colonies were grown overnight in 3 ml LB/ampicillin. After overnight growth at 37°C, with a 200 rpm circular agitation, the plasmids were miniprepped according to classical protocols. The cDNA concentration was evaluated after running on an Ethidium Bromide-stained 1% agarose gel, and 600 ng were sequenced using a 16-capillar Applied Biosystem sequencing machine, using an Applied Biosystem sequencing kit.
Quantification of the signals and statistical analysis
Each signal was quantified by densitometry using the Scion software [56]. Data were entered in an Excel table encompassing 2304 rows and one column for each signal obtained in a hybridization experiment. Data were normalized by reference to a maximum signal intensity fixed at 120 (arbitrary units). They were then grouped into six categories corresponding to the averages of the hybridization experiments, ETP (Early Term Placenta) ETP48 h (Early Term Placenta maintained 48 h at 2% O2), FTP (Control Full Term Placenta), PE (Preeclampsia), R (Intra-Uterine Growth retardation), PER (Preeclampsia and Intra Uterine Growth Retardation). Data were then clustered in a hierarchical tree using the software developed by Eisen and coworkers, Cluster and Treeview, available at [22,57]. The program Cluster was used after the following data preparation: median centering and normalization of genes and arrays, before launching the "complete clustering" procedure. The distribution of groups of genes expressed in only one situation, was analyzed in relation to expected distribution in the normoxic/hypoxic plates of clones (plates 1 to 9 resulted from the normoxia – hypoxia subtraction, while plates 10 to 24 resulted from hypoxia – normoxia subtraction. The distributions could therefore be tested by a Chi2 analysis.
A simple mathematical method was implemented for profiling the CpG islands present in the promoters of the genes identified. Briefly, all the CpG positions were identified in the 5 kb upstream of transcription initiation start site for each gene, and the distance between two consecutive CpG was computed. Then, a mobile average was computed for sliding windows corresponding to ten consecutive CpG islands. A CpG density was estimated at each position by dividing 1000 by the average of each sliding window, resulting in the number of CpG per kilobases of DNA.
To study statistically the chromosomal distribution of the highly expressed subset of placental genes from the macroarrays, we evaluated the size of the minimal statistically significant interval expected to contain 2, 3, 4, 5, 6, 7 or 8 genes. This was calculated using a Binomial law with 2 to 8 successes amongst 276 trials, with a probability of success being estimated as the size of the interval divided by the genome size (3000 megabases; in this study, the gene distribution was supposed to be similar for each chromosome, which is an approximation since some chromosomes are gene-rich and some others are gene-poor). The probabilities were corrected by the Bonferroni correction, available at [58] for multiple testing, assuming 276 independent tests. Using this procedure, the minimal significant genomic interval sizes could be estimated at 0.21, 1.18, 3.05, 5.70, 9.05, 13.0 and 17.30 Mb for 2, 3, 4, 5, 6, 7, or 8 genes respectively (The complete set of data is available upon request).
Authors' contributions
DV coordinated the program and constructed the SSH libraries, carried out several hybridizations, read the films, analyzed the results at the statistical and biological level and wrote the draft of the article. FM and AGD purify the placental villi from the decidue, put them in culture, purify the total and polyA+ RNA, participated in the SSH and performed a large part of the hybridization experiments. TMM read the films independently and managed the medical files of the patients. BR prepared part of the cDNA from the mRNA samples, as well as HJ. RR took care of the cultures in hypoxia enabling to prepare the cDNA from early villi. STC and FQ contributed to the fine characterization of the membranes. GM and VS helped in the characterization of the membranes by the gift of cDNA samples from JEG-3 cell cultures (results not included in the present paper). FP was in charge of the robotic spotting enabling to construct the membranes. JLD carried out the primary hypoxic culture that were used for the SSH experiments, VR and BC are both clinicians that were in charge of collecting the patients at the Institut de Puériculture and St Antoine hospital, respectively. FF helped in the redaction of the paper.
Acknowledgements
The sequencing service of the Cochin platform is greatly acknowledged. This work was funded by INSERM.
==== Refs
Nilsson E Salonen Ros H Cnattingius S Lichtenstein P The importance of genetic and environmental effects for pre-eclampsia and gestational hypertension: a family study Bjog 2004 111 200 206 14961879
Esplin MS Fausett MB Fraser A Kerber R Mineau G Carrillo J Varner MW Paternal and maternal components of the predisposition to preeclampsia N Engl J Med 2001 344 867 872 11259719 10.1056/NEJM200103223441201
Zusterzeel PL te Morsche R Raijmakers MT Roes EM Peters WH Steegers EA Paternal contribution to the risk for pre-eclampsia J Med Genet 2002 39 44 45 11826024 10.1136/jmg.39.1.44
Pijnenborg R Robertson WB Brosens I The arterial migration of trophoblast in the uterus of the golden hamster, Mesocricetus auratus J Reprod Fertil 1974 40 269 280 4372344
Robertson WB Uteroplacental vasculature J Clin Pathol Suppl (R Coll Pathol) 1976 9 17 1085779
Pijnenborg R Dixon G Robertson WB Brosens I Trophoblastic invasion of human decidua from 8 to 18 weeks of pregnancy Placenta 1980 1 3 19 7443635
Brosens JJ Pijnenborg R Brosens IA The myometrial junctional zone spiral arteries in normal and abnormal pregnancies: a review of the literature Am J Obstet Gynecol 2002 187 1416 1423 12439541 10.1067/mob.2002.127305
Kaufmann P Black S Huppertz B Endovascular trophoblast invasion: implications for the pathogenesis of intrauterine growth retardation and preeclampsia Biol Reprod 2003 69 1 7 12620937 10.1095/biolreprod.102.014977
Damsky CH Fisher SJ Trophoblast pseudo-vasculogenesis: faking it with endothelial adhesion receptors Curr Opin Cell Biol 1998 10 660 666 9818178 10.1016/S0955-0674(98)80043-4
Dekker GA Robillard PY Hulsey TC Immune maladaptation in the etiology of preeclampsia: a review of corroborative epidemiologic studies Obstet Gynecol Surv 1998 53 377 382 9618714 10.1097/00006254-199806000-00023
Granger JP Alexander BT Llinas MT Bennett WA Khalil RA Pathophysiology of preeclampsia: linking placental ischemia/hypoxia with microvascular dysfunction Microcirculation 2002 9 147 160 12080413 10.1038/sj.mn.7800137
Kingdom JC Kaufmann P Oxygen and placental vascular development Adv Exp Med Biol 1999 474 259 275 10635006
Goldman-Wohl D Yagel S Regulation of trophoblast invasion: from normal implantation to pre-eclampsia Mol Cell Endocrinol 2002 187 233 238 11988332 10.1016/S0303-7207(01)00687-6
Challier JC Uzan S [The human placenta and its pathologies: focus on oxygen] Med Sci (Paris) 2003 19 1111 1120 14648482
Barker DJ Eriksson JG Forsen T Osmond C Fetal origins of adult disease: strength of effects and biological basis Int J Epidemiol 2002 31 1235 1239 12540728 10.1093/ije/31.6.1235
Barker DJ The developmental origins of adult disease Eur J Epidemiol 2003 18 733 736 12974544 10.1023/A:1025388901248
Wang JX Knottnerus AM Schuit G Norman RJ Chan A Dekker GA Surgically obtained sperm, and risk of gestational hypertension and pre-eclampsia Lancet 2002 359 673 674 11879865 10.1016/S0140-6736(02)07804-2
Mondon F Mignot TM Rebourcet R Jammes H Danan JL Ferre F Vaiman D Profiling of oxygen-modulated gene expression in early human placenta by systematic sequencing of suppressive subtractive hybridization products Physiol Genomics 2005
Nielsen FC Nielsen J Christiansen J A family of IGF-II mRNA binding proteins (IMP) involved in RNA trafficking Scand J Clin Lab Invest Suppl 2001 234 93 99 11713986 10.1080/003655101317095473
Constancia M Hemberger M Hughes J Dean W Ferguson-Smith A Fundele R Stewart F Kelsey G Fowden A Sibley C Reik W Placental-specific IGF-II is a major modulator of placental and fetal growth Nature 2002 417 945 948 12087403 10.1038/nature00819
Liao B Hu Y Herrick DJ Brewer G The RNA-binding protein IMP-3 is a translational activator of insulin-like growth factor II leader-3 mRNA during proliferation of human K562 leukemia cells J Biol Chem 2005
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 10.1073/pnas.95.25.14863
Merviel P Muller F Guibourdenche J Berkane N Gaudet R Breart G Uzan S Correlations between serum assays of human chorionic gonadotrophin (hCG) and human placental lactogen (hPL) and pre-eclampsia or intrauterine growth restriction (IUGR) among nulliparas younger than 38 years Eur J Obstet Gynecol Reprod Biol 2001 95 59 67 11267722 10.1016/S0301-2115(00)00370-5
Bersinger NA Groome N Muttukrishna S Pregnancy-associated and placental proteins in the placental tissue of normal pregnant women and patients with pre-eclampsia at term Eur J Endocrinol 2002 147 785 793 12457454 10.1530/eje.0.1470785
Semenza GL Agani F Booth G Forsythe J Iyer N Jiang BH Leung S Roe R Wiener C Yu A Structural and functional analysis of hypoxia-inducible factor 1 Kidney Int 1997 51 553 555 9027737
NCBI National Center for Biotechnology information [http://www.ncbi.nlm.nih.gov/mapview/]
Iccare Iccare database for comparative mapping [http://genopole.toulouse.inra.fr/Iccare/]
Hayes H Chromosome painting with human chromosome-specific DNA libraries reveals the extent and distribution of conserved segments in bovine chromosomes Cytogenet Cell Genet 1995 71 168 174 7656590
Schibler L Vaiman D Oustry A Giraud-Delville C Cribiu EP Comparative gene mapping: a fine-scale survey of chromosome rearrangements between ruminants and humans Genome Res 1998 8 901 915 9750190
Annaleena H Tiina T Sanna-Kaisa H Leea KN Seppo H Seppo YH Tumor suppressor and growth regulatory genes are overexpressed in severe early-onset preeclampsia--an array study on case-specific human preeclamptic placental tissue Acta Obstet Gynecol Scand 2005 84 679 689 15954879 10.1111/j.0001-6349.2005.00814.x
Bersinger NA Odegard RA Second- and third-trimester serum levels of placental proteins in preeclampsia and small-for-gestational age pregnancies Acta Obstet Gynecol Scand 2004 83 37 45 14678084 10.1111/j.1600-0412.2004.00277.x
Loregger T Pollheimer J Knofler M Regulatory transcription factors controlling function and differentiation of human trophoblast--a review Placenta 2003 24 Suppl A S104 10 12842421 10.1053/plac.2002.0929
Arnold LL Doherty TM Flor AW Simon JA Chou JY Chan WY Mansfield BC Pregnancy-specific glycoprotein gene expression in recurrent aborters: a potential correlation to interleukin-10 expression Am J Reprod Immunol 1999 41 174 182 10326620
Snyder SK Wessner DH Wessells JL Waterhouse RM Wahl LM Zimmermann W Dveksler GS Pregnancy-specific glycoproteins function as immunomodulators by inducing secretion of IL-10, IL-6 and TGF-beta1 by human monocytes Am J Reprod Immunol 2001 45 205 216 11327547 10.1111/j.8755-8920.2001.450403.x
Motran CC Diaz FL Gruppi A Slavin D Chatton B Bocco JL Human pregnancy-specific glycoprotein 1a (PSG1a) induces alternative activation in human and mouse monocytes and suppresses the accessory cell-dependent T cell proliferation J Leukoc Biol 2002 72 512 521 12223519
Tsoi SC Cale JM Bird IM Kay HH cDNA microarray analysis of gene expression profiles in human placenta: up-regulation of the transcript encoding muscle subunit of glycogen phosphorylase in preeclampsia J Soc Gynecol Investig 2003 10 496 502 14662163 10.1016/S1071-5576(03)00154-0
Yoon JC Chickering TW Rosen ED Dussault B Qin Y Soukas A Friedman JM Holmes WE Spiegelman BM Peroxisome proliferator-activated receptor gamma target gene encoding a novel angiopoietin-related protein associated with adipose differentiation Mol Cell Biol 2000 20 5343 5349 10866690 10.1128/MCB.20.14.5343-5349.2000
Yoshida K Shimizugawa T Ono M Furukawa H Angiopoietin-like protein 4 is a potent hyperlipidemia-inducing factor in mice and inhibitor of lipoprotein lipase J Lipid Res 2002 43 1770 1772 12401877 10.1194/jlr.C200010-JLR200
Kersten S Mandard S Tan NS Escher P Metzger D Chambon P Gonzalez FJ Desvergne B Wahli W Characterization of the fasting-induced adipose factor FIAF, a novel peroxisome proliferator-activated receptor target gene J Biol Chem 2000 275 28488 28493 10862772 10.1074/jbc.M004029200
Ripoche MA Kress C Poirier F Dandolo L Deletion of the H19 transcription unit reveals the existence of a putative imprinting control element Genes Dev 1997 11 1596 1604 9203585
Zaina S Pettersson L Normal size of the lumen of the aorta in dwarf mice lacking IGF-II Growth Horm IGF Res 2001 11 298 302 11735248 10.1054/ghir.2001.0241
Zaina S Pettersson L Ahren B Branen L Hassan AB Lindholm M Mattsson R Thyberg J Nilsson J Insulin-like growth factor II plays a central role in atherosclerosis in a mouse model J Biol Chem 2002 277 4505 4511 11726660 10.1074/jbc.M108061200
Georgieff MK Wewerka SW Nelson CA Deregnier RA Iron status at 9 months of infants with low iron stores at birth J Pediatr 2002 141 405 409 12219063 10.1067/mpd.2002.127090
Rosen FS Alper CA Pensky J Klemperer MR Donaldson VH Genetically determined heterogeneity of the C1 esterase inhibitor in patients with hereditary angioneurotic edema J Clin Invest 1971 50 2143 2149 4107267
Tosi M Molecular genetics of C1 inhibitor Immunobiology 1998 199 358 365 9777419
Blanch A Roche O Lopez-Granados E Fontan G Lopez-Trascasa M Detection of C1 inhibitor (SERPING1/C1NH) mutations in exon 8 in patients with hereditary angioedema: evidence for 10 novel mutations Hum Mutat 2002 20 405 406 12402344 10.1002/humu.9073
Kalmar L Bors A Farkas H Vas S Fandl B Varga L Fust G Tordai A Mutation screening of the C1 inhibitor gene among Hungarian patients with hereditary angioedema Hum Mutat 2003 22 498 14635117 10.1002/humu.9202
Kanayama N Takahashi K Matsuura T Sugimura M Kobayashi T Moniwa N Tomita M Nakayama K Deficiency in p57Kip2 expression induces preeclampsia-like symptoms in mice Mol Hum Reprod 2002 8 1129 1135 12468647 10.1093/molehr/8.12.1129
Lercher MJ Urrutia AO Hurst LD Clustering of housekeeping genes provides a unified model of gene order in the human genome Nat Genet 2002 31 180 183 11992122 10.1038/ng887
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
Cosseddu GM Perez-Enciso M Fellous M Vaiman D Interspecific chromosome-wide transcription profiles reveal the existence of mammalian-specific and species-specific chromosome domains J Mol Evol 2004 59 317 28 15553087 10.1007/s00239-004-2627-6
Cremer T Cremer C Chromosome territories, nuclear architecture and gene regulation in mammalian cells Nat Rev Genet 2001 2 292 301 11283701 10.1038/35066075
Diatchenko L Lau YF Campbell AP Chenchik A Moqadam F Huang B Lukyanov S Lukyanov K Gurskaya N Sverdlov ED Siebert PD Suppression subtractive hybridization: a method for generating differentially regulated or tissue-specific cDNA probes and libraries Proc Natl Acad Sci U S A 1996 93 6025 6030 8650213 10.1073/pnas.93.12.6025
Chomczynski P A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples Biotechniques 1993 15 532 4, 536-7 7692896
Sambrook J Fritsch EF Maniatis T Molecular Cloning: A Laboratory Manual 1989 , Cold Spring Harbor Laboratory Press
Scion Scion acquisition and analysis software [http://www.scioncorp.com/frames/fr_scion_products.htm]
Eisen Eisen software for cluster analysis and vizualisation [http://rana.lbl.gov/EisenSoftware.htm]
Bonferroni Bonferroni correction for multiple testing [http://home.clara.net/sisa/bonfer.htm]
|
16129025
|
PMC1236921
|
CC BY
|
2021-01-04 16:32:48
|
no
|
BMC Genomics. 2005 Aug 29; 6:111
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-111
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1171615015710.1186/1471-2164-6-117Research ArticleComprehensive search for intra- and inter-specific sequence polymorphisms among coding envelope genes of retroviral origin found in the human genome: genes and pseudogenes de Parseval Nathalie [email protected] Gora [email protected] Sandra [email protected] François [email protected] Alexandre [email protected] Fumihiko [email protected] Thierry [email protected] UMR 8122 CNRS, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif Cedex, France2 Centre National de Génotypage, 2, rue Gaston Crémieux, Évry, France2005 9 9 2005 6 117 117 16 3 2005 9 9 2005 Copyright © 2005 de Parseval et al; licensee BioMed Central Ltd.2005de Parseval 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 human genome carries a high load of proviral-like sequences, called Human Endogenous Retroviruses (HERVs), which are the genomic traces of ancient infections by active retroviruses. These elements are in most cases defective, but open reading frames can still be found for the retroviral envelope gene, with sixteen such genes identified so far. Several of them are conserved during primate evolution, having possibly been co-opted by their host for a physiological role.
Results
To characterize further their status, we presently sequenced 12 of these genes from a panel of 91 Caucasian individuals. Genomic analyses reveal strong sequence conservation (only two non synonymous Single Nucleotide Polymorphisms [SNPs]) for the two HERV-W and HERV-FRD envelope genes, i.e. for the two genes specifically expressed in the placenta and possibly involved in syncytiotrophoblast formation. We further show – using an ex vivo fusion assay for each allelic form – that none of these SNPs impairs the fusogenic function. The other envelope proteins disclose variable polymorphisms, with the occurrence of a stop codon and/or frameshift for most – but not all – of them. Moreover, the sequence conservation analysis of the orthologous genes that can be found in primates shows that three env genes have been maintained in a fully coding state throughout evolution including envW and envFRD.
Conclusion
Altogether, the present study strongly suggests that some but not all envelope encoding sequences are bona fide genes. It also provides new tools to elucidate the possible role of endogenous envelope proteins as susceptibility factors in a number of pathologies where HERVs have been suspected to be involved.
==== Body
Background
A large fraction (8%) of the human genome contains elements of retroviral origin, with thousands of sequences closely related to the integrated proviral form of infectious retroviruses with the canonical gag, prt, pol and env genes [1]. These elements, named human endogenous retroviruses (HERV), are most probably the proviral remnants of ancestral germline infections by active retroviruses, which have thereafter been transmitted in a Mendelian manner. HERVs have been grouped according to sequence homologies into more than 80 distinct families, each containing a few to several hundreds elements (reviewed in [2-4], see [5] for their classification). Most HERV genes are non-coding, due to either terminating mutations or deletions, but we have characterized 16 human endogenous env genes that have retained a coding capacity among the 30,000 endogenous proviral elements of the human genome [6]. The analysis of their transcriptome in healthy human tissues has revealed that three of them are highly expressed in the placenta, namely the erv3/HERV-R, the HERV-W and the HERV-FRD env genes [6].
Phylogenetic as well as functional analyses have revealed strong similarities between HERVs and the present-day infectious retroviruses, suggesting a common history and shared ancestors. Accordingly, it has been proposed that HERVs may still possess some of the functions of infectious retroviruses and as such have pathogenic effects, provided that they are transcriptionally active. Conversely, it is also plausible that HERV proteins may have been co-opted by the host for its benefit. Along this line, it has been proposed that the HERV envelope proteins could play a role in several processes including i) protection against infection by present-day retroviruses through receptor interference [7], ii) protection of the fetus against the maternal immune system via an immunosuppressive domain located in the envelope transmembrane (TM) subunit [8,9], and iii) placenta morphogenesis through fusogenic effects allowing differentiation of cytotrophoblastic cells into the syncytiotrophoblast [10-12]. In accordance with a symbiotic role for HERVs, it has recently been shown that the HERV-W and HERV-FRD envelope gene products are highly fusogenic glycoproteins that are specifically expressed in the placenta and can mediate cell-cell fusion ex vivo [12,13]. Involvement of HERV proteins in physiological processes, however, remains a debated issue, and definite evidence is still lacking. Because selection pressure on a functional gene should result in a limited mutation rate, the survey of single nucleotide polymorphisms (SNPs) among the human population is a way to evaluate functional constraints on these genes. Using this approach, we had previously demonstrated that one postulated candidate for a role in placentation, namely the highly-expressed erv3/HERV-R envelope gene carries a homozygous stop mutation resulting in a severe protein truncation in 1% of individuals of caucasian origin, which strongly suggests that it is not necessary for any fundamental placental function [14]. The unexpectedly low number of still coding envelope genes present in the human genome [6] now allows a comprehensive analysis of such genes to be performed, in order to assess their possible physiological and/or physiopathological role. Here, we analysed the SNP level of the 12 coding env genes present in the human genome that could be characterized by this approach, together with their conservation among the orthologous genes that can be identified in primates. The two series of data are consistent with a role beneficial to the host for some of the genes, whereas others are likely to be subjected to progressive inactivation. In both cases, the identified SNPs should be useful tools to evaluate the possible role of these genes as "susceptibility genes" in several human pathologies where HERVs have been suspected to be involved.
Results
Structure and PCR-amplification of the fully coding HERV envelope genes
Retroviral envelope genes are 2 kb-long sequences with no introns, that are located in the 3' domain of proviral elements (see Figure 1A for the genomic structure of a provirus and Figure 1B and 1C for the description of the envelope gene and its product). Endogenous retroviruses being in most cases highly reiterated elements (see Table 1), the 3' PCR primers for env gene amplification had to be placed downstream of the provirus end to specifically amplify the family member of interest, whereas the 5' primer was placed upstream of the env MET initiation codon (see Figure 1A). Among the 16 coding env genes that we had previously identified [6], 12 could be subjected to a systematic search for SNPs, including 3 out of the 6 HERV-K(HML-2) coding env genes (K1, K2, K4), 2 out of the 3 HERV-H coding env genes (H1 and H3), and the coding env genes of the F(c)1, F(c)2, T, W, R, R(b) and FRD families. Three HERV-K(HML-2) env genes could not be PCR amplified: the first one (envK3) is located in the centromeric region of chromosome 19, and its provirus is surrounded by stretches of repeated sequences, thus precluding the use of specific primers; the other two (envK5 and envK6) belong to proviruses present only in a fraction of the caucasian population [15], with a low allele frequency (0.19 and 0.04, respectively), thus precluding a statistically significant SNP study to be performed (unless a pre-selected population was used). It is noteworthy that proviruses K1, K2 and K4 have also been recently demonstrated to be polymorphic, since they exist in some individuals as a solo Long Terminal Repeat (LTR), devoid of internal sequences [16]. Yet, the allele frequencies of the complete provirus forms are high (0.72, 0.97 and 0.89, respectively), thus allowing an SNP study. Finally, sequence data for one of the 3 HERV-H coding env genes studied (i.e. envH2) yielded multiple sequence profiles (possibly due to the parasitic amplification of another member of the HERV-H family) and could not be analyzed further. The list of the 12 fully coding env genes that could finally be analyzed, together with the associated amplification primers used, is given in Table 1.
Figure 1 Schematic representation of the proviral form of a retrovirus and its env gene products. A, Genomic structure of a provirus, with the four canonical retroviral genes gag (encoding the virion core proteins), prt (encoding the protease), pol (encoding the reverse transcriptase, RNAseH, integrase), and env (encoding the viral membrane protein). The position of the primers designed to amplify the coding env gene is indicated. B, Linear representation of a retroviral envelope protein and its functional domains. The env gene encodes a polypeptide which is cleaved into two proteins, the surface protein (SU), which is involved in receptor recognition, and the transmembrane subunit (TM), which anchors the whole envelope complex to the membrane and is directly responsible for cell membrane fusion and virus entry. The TM subunit contains three functional domains, the fusion peptide, the immunosuppressive domain and the transmembrane domain. C, Schematic structure of the envelope gene products. C-C: disulfide bound.
Table 1 The coding envelope genes of the human genome studied for their SNPs.
Gene name Bibliographic gene name Family name (repbase identifier) Approximate number of elements Genomic localization Amplification primers sequence (5'-3')
envK1 - HERV-K(HML-2) (HERVK) 50–100 Chr12: 57008431–57010527 (-) F: GGGAAATAGGGAAGGTGATA
R: ACATCCCTAACGCTTTAAG
envK2 HML-2.HOM, HERV-K108 idem idem Chr7: 4367317–4369416 (-) F: GAGGTTTTGCTTGTGTTTCA
R: TTAGGCTTTCGGGACTTCAA
envK4 HERV-K109 idem idem Chr6: 78423172–78425268 (-) F: GGGAAATAGGGAAGGTGATA
R: GGGTAGTATCAGTCGGGATG
envF(c)1 - HERV-F(c)1 (-) 1 Chr:X: 95874118–95875872 (+) F: GCACCGACTCAGCACGAC
R: GCCTTGGCAACTAAACCATTC
envF(c)2 - HERV-F(c)2 (-) 15 Chr7:152498167–152499936 (-) F: GAAGGCACCTACACAACATC
R: GACACTTAATAGTTGCGACA
envH1 envH/p62, H19 HERV-H (HERVH) 1000 Chr2: 166767244–166768998 (-) F: ATGCCCTACTCTTGTTTACAC
R: AAATCTGGCAAACTACAAGC
envH3 envH/p59 idem idem Chr2: 155931277–155932944 (+) F: TTTCTTCAAGCCATCACAGC
R: ACCCCATGTTCTAGTCTTCC
envT - HERV-T (HERVS71) 50 Chr19: 20341241–20343121 (+) F: TTGGATTCATCACTCCCA
R: CTGAAGGGAGTTCCTCCTAGG
envW Syncytin 1 HERV-W (HERV17) 200 Chr7: 91710108–91711724 (-) F: AACAACCAGGAGGAAAGTAA
R: CTGATCAAGTCGCAAAGC
envR erv3 HERV-R (HERV3) 100 Chr7: 63863079–63865094 (-) F: GGTTAGAAATCTGAAGTCC
R: AAAGTCAATGACAGATGCGG
envR(b) - HERV-R(b) (PABL_B) 50 Chr3: 16786814–16788358 (+) F: GCTAAGCACCAGTTCAGCACTG
R: TGTTTTGGGACACCACGAAT
envFRD Syncytin 2 HERV-FRD (MER50) 3000 Chr:6: 11211913–11213529 (-) F: CTTGTACACCACCAGGAGTTCC
R: TTTGAGCAAGGGTGATTCAT
SNP of the human coding envelope genes and haplotype analysis
The 12 HERV coding env genes were PCR-amplified for 91 healthy Caucasian individuals, and each PCR product was directly sequenced without cloning. The identified SNPs are positioned on the protein sequences in Figure 2, and the number and nature of the variations (synonymous [i.e. silent] versus nonsynonymous [i.e. leading to an amino acid change] substitutions, stop or insertion/deletion [indel] mutations) are given in Figure 3. The complete list and detailed features – including frequencies – of the identified SNPs is available in additional data file 1. As illustrated in Figure 2 and 3, the number of SNPs for the 12 env genes varies significantly depending on the gene, but their distribution along the envelope sequences seems to be random. The number of SNPs per kb varies from 1 (for envW) to 10 (for envT). Studies on SNPs in intragenic regions of the human genome disclose average densities varying from 0.8 to 5 SNPs per kb [17,18]. These studies also provide an estimate of the average ratio of synonymous versus non-synonymous SNPs, which is close to 1 in coding regions, whereas it is close to 0.5 in pseudogenes [17,18]. For the env genes, provided that the SNP number is high enough to validate this ratio (thus excluding envFRD, envW, envF(c)1, envH1 and envH3), it ranges from 0.11 for envR(b) to 0.66 for envT, close to the pseudogene ratio.
Figure 2 Localization of the SNPs on the envelope genes. Characteristic domains of the envelope proteins are depicted as in Figure 1B. The gray frames at the end of H3, R and F(c)2 genes represent short open reading frames present downstream of the stop codon. Mutations are represented along the protein sequence, with the number corresponding to the amino acid affected by the SNP. Synonymous mutations are indicated above the protein frame. Non-synonymous (with stop mutations indicated) and indel mutations (with frame shifting mutations underlined) are indicated below the protein frame.
Figure 3 Number of variations identified in the HERV coding envelope genes. The SNP numbers are indicated for each env gene, with the number of synonymous SNPs, nonsynonymous SNPs except stops, and stops or indel mutations indicated. The env genes are ordered according to their date of entry in the primate lineage (see [4]), indicated in million years (Myr).
Based on the number of nonsynonymous and indel mutations, a hierarchy among the endogenous coding env genes can be established, with the envW, envFRD and envF(c)1 genes being the most conserved (see also [19]), and the envK1, envK2, envT, envR, envR(b), and envF(c)2 genes being affected by numerous mutations, including mutations resulting in truncation of the protein due to frameshifting or generation of stop codons. There is no correlation between the number of SNPs and the "age" of the corresponding gene in the primate lineage. This is clearly illustrated in Figure 3, where the env genes have been ordered according to the date of entry of each corresponding provirus into the host genome as previously determined via an analysis of the orthologous loci throughout evolution (reviewed in [4]). For instance, the env genes of the HERV-K(HML-2) family are human-specific, i.e. are present in the genome of primate since less than five million years (Myrs), whereas envR(b) and envFRD have entered the genome of the common ancestor of Old World and New World monkeys more than 40 Myrs ago. For the envFRD gene, which is among the "oldest" env genes, only 4 SNPs are found, whereas for the "recent" env genes of the HERV-K(HML-2) family, the SNP number can be as high as 16. Although the lack of correlation between the "age" of the genes in the primate lineage and the numbers of SNPs is not unexpected taking into account the occurrence of "bottlenecks" giving rise to founder effects during the evolution of the human population, what remains surprising is the important variability of the number and "severity" of SNPs among the env genes. This should be a strong indication for a differential selection pressure exerted on these genes (see below).
A further characterization of the SNPs, including genotype distribution, haplotype frequency, and linkage disequilibrium was performed (additional data files 1, 2 and 3). The genotype distributions are compatible with a Hardy-Weinberg equilibrium, except for some positions on the HERV-K1 (HERV-K1_44, HERV-K1_133, HERV-K1_403, HERV-K1_482, HERV-K1_651, HERV-K1_673, HERV-K1_2144), HERV-K4 (HERV-K4_292, HERV-K4_369, HERV-K4_382, HERV-K4_586) and HERV-F(c)1 (HERV-F(c)1_122, HERV-F(c)1_226, HERV-F(c)1_235) env genes, consistent with recent integration of these elements in the primate lineage (see Figure 3). Haplotype frequencies were estimated for each gene based on the Expectation-Maximization (EM) algorithm [20] for haplotypes with frequency estimates >1%. The results are summarized in additional data file 2. The three most frequent haplotypes for each env gene represent >80% of all the haplotypes, suggesting that these regions have a low recombination rate. A linkage disequilibrium (LD) plot was generated (additional data file 3) with pairwise LDs measured between each pair of polymorphisms using the D and D' methods (see materials and methods). As expected, the majority of high LD values occur within the env genes, and the LDs calculated between SNPs of different HERV env genes is low. Low LDs were obtained as well for env genes located on the same chromosome, i.e. for HERV-F(c)2, HERV-K2, HERV-R, HERV-W (on chromosome 7) and for HERV-H1 and HERV-H3 (on chromosome 2; 10 Mb apart). These observations suggest an independent evolution for each env gene.
Among all the coding env genes, envW and envFRD are the only two genes with a clearly identified functional property, i.e. the capacity to generate cell-cell fusion [10,12]. This property – most probably associated with its role in placentation, see the Background section – was used to characterize further the consequences of the identified SNPs on the fusogenic function of the encoded proteins. To do so, we PCR-amplified the genomic DNAs of individuals carrying the corresponding SNP alleles, with primers allowing the cloning of the env genes in appropriate expression vectors. The fusogenic function was then assayed as in [12], using two different cell lines for fusion (Figure 4). As illustrated in the figure, no difference can be observed between the four haplotypes of each of the env genes tested. This, together with the low SNP level for the two genes, is a strong indication for selection of a "function" associated with the corresponding proteins.
Figure 4 Effects of the non synonymous SNPs identified in envFRD and envW on their fusogenic function. A, Construction of the env-expression vectors and rationale of the fusion assay. Each of the env allelic forms were PCR amplified from genomic DNA and cloned into the phCMV expression vector. Cells were transfected with the env-expression vectors and stained with May-Grünwald and Giemsa solutions. B, Cell-cell fusion assay for the allelic forms of envFRD (upper panel) and envW (lower panel), using two cells types (human 293T and TE671 cells). The fusion index represents the percentage of fusion events in the transfected cell populations as evidenced by syncytia formation, and is quantitated as in [12]. The control corresponds to transfection of an expression vector without an env gene. The most frequent haplotypes correspond to FRD359A/367T and W138R/307S.
Interspecific sequence conservation
The data regarding the orthologous genes that can be found in primates for each of the presently studied human env genes are given in Figure 5. Some of the orthologous primate env genes had been previously cloned and sequenced ([19] for envW, [21] for envFRD, [22] for envR and [23] for envF(c)1). Others have been PCR-amplified and tested using a direct coupled in vitro transcription/translation assay to determine their coding status ([24] for the 3 envH and this study for envT, envF(c)2 and envR(b), see Materials and Methods). As illustrated in the figure, a first important outcome is that only three env genes have been maintained in a fully coding state throughout evolution, i.e. envFRD (7/7 lineages), envW (5/5 lineages) and envR (5/5 lineages) which, interestingly, correspond to the three env genes highly expressed in the placenta. Secondly, as observed for the human SNP analysis, there is no correlation between the "age" of the gene in the primate lineage and its coding status: the latter three env genes are among the "oldest" ones, whereas three other env genes (envH3, envT and envF(c)1) are fully coding in only one non-human lineage, and the others are only coding in the human lineage.
Figure 5 Conservation of the coding envelopes in primate species. Hum.: human, Chimp.: chimpanzee, Gor.: gorilla, Orang.: orangutan, Gib.: gibbon, OWM: Old World Monkeys, NWM: New World Monkeys. The boxes indicate presence of the env gene, with its coding status illustrated with a color code.
Discussion
The present investigation of the fully coding human env genes, including the human SNP search and the analysis of the coding status of the identified primate orthologs, pinpoints two of these genes – namely envFRD and envW – that disclose the characteristic features of a gene subjected to a functional constraint, i.e. low polymorphism and maintenance of an open reading frame during evolution. Interestingly, these two genes are highly – and specifically – expressed in the placenta, and possess a well-characterized fusogenic function which led to the proposal that they are bona fide genes that have been co-opted by the host for a physiological function related to placenta physiology [10,12,13,19]. Among the other genes, envR is of interest since as with the two former genes it is highly expressed in the placenta and has maintained its fully coding status in all species since its entry into the primate lineage. Yet, the SNP analysis discloses a severe polymorphism – including a premature stop codon – indicating that the preservation of an open reading frame during evolution should not be considered as a sufficient criterion to assign a biological function. The other genes are not conserved in a fully coding state in all the primate branches where they are present and/or disclose a severe polymorphism (with in several cases occurrence of a premature stop codon for some allelic forms).
Another hallmark of the presence of a functional constraint on a gene is a low nonsynonymous/synonymous substitution ratio (Ka/Ks) for orthologous genes found in other species. A bona fide gene with a cellular function should be under purifying selection, which prevents deleterious nonsynonymous substitutions from being fixed and usually does not affect synonymous substitutions, leading to Ka/Ks ratios <1, whereas a gene under genetic drift (e.g. a pseudogene) has a Ka/Ks ratio close to unity (reviewed in [25]). Such an analysis has already been performed for the envFRD and the envW genes. The mean ratio for all pairwise comparisons was 0.37 for envFRD [21], demonstrating the existence of a selective pressure. For envW, the corresponding ratio was 0.8 [19], precluding definite conclusions (yet, a subsequent study on envW identified a region of the gene with a lower ratio compatible with a functional constraint specifically exerted on that domain [26]). We have calculated the Ka/Ks ratios for the other env genes, when nucleotide sequences of their orthologs were available (data not shown). These ratios were found to be heterogeneous for a given env gene, with values ranging from 0.23 to 1.17, again precluding any definite conclusion to be drawn.
According to the present analysis of intra- and interspecific variability of the env genes, one is led to conclude that most probably only two among the twelve studied env genes are likely to be involved in an essential human physiological function, whereas the others would be on their way to conversion into pseudogenes. The presence of a reading frame still open in human for the latter genes may appear intriguing, but one has to keep in mind that they belong to multicopy HERV families, and as such one of the element could have remained open just by chance, without any purifying selection, since even under completely neutral drift it takes time for a sufficient number of mutations to transform a gene into a pseudogene. Along this line, it is of interest to mention the study by Zhang and Webb on the primate V1R pheromone receptor genes, for which there were approximately 140 copies in the genome of the common ancestor of Old World monkeys and hominoids, whereas the human genome has only five V1R genes that retained an ORF. Examination of the orthologous genes in primates showed that none of the five genes kept an intact ORF in all of the apes. Furthermore, for the orthologous sequences with an intact ORF, Ka/Ks ratios were close to unity. The intraspecific variation of these five human genes was also assessed, and for two of them an allelic form generating a premature stop codon was found. Altogether, the authors concluded that there were no functional constraints on these genes since before the separation of hominoids and Old World Monkeys (approximately 23 Myrs ago) and that they were in the process of pseudogeneization in those primate species [27]. Another possible explanation for the "neutral" conservation of an open reading frame for an HERV env gene without any selection pressure from the host could be related to the relatively autonomous status of these parasitic elements, and be associated with the persistence of active retroviral elements responsible for the maintenance – by a reiterated infection process – of some of the HERV families (e.g. [23,28]).
In any case, it appears clearly that conservation of an env gene with a coding status cannot be taken as the sole criterion for a possible function to the benefit of the host, with only the envW and envFRD genes emerging from the present study as possible bona fide genes. Yet, this does not mean that the other env genes cannot have any effect in humans. Indeed, the present analysis only indicates that they are not under stringent purifying selection, in terms of evolution, but they still could be involved in pathologies – such as tumors or auto-immune diseases – not deleterious to the species because they occur late in the life of the individuals. One should keep in mind that endogenous retroviruses originate from bona fide retroviruses, and as such might have conserved some of the pathological potency of their progenitors. In this respect, the identified SNPs should be essential tools to determine if this is indeed the case, via an analysis of their distribution among selected groups of individuals with a definite pathology. Along this line, the present data on the envT gene are of special interest. Indeed, this gene is the only non-placental env gene found to be highly expressed in a human tissue – the thyroid – of healthy individuals [6], and the high level of polymorphism of the gene shown in this report together with its lack of conservation in primates suggest that it is not involved in any essential physiological process and thus not subjected to purifying selection. Thanks to the identified SNPs, it can now be tested whether this expressed gene is involved in a pathological process in humans, among which thyroid tumors could be select candidates for a systematic search.
Conclusion
The systematic SNP search on fully coding human endogenous envelope genes, combined with an analysis of the sequence conservation among the orthologs that can be identified among primates revealed that two genes (envW and envFRD) can be considered as bona fide genes, and identified polymorphisms – to a variable extent – in the other genes. The data are consistent with a physiological role for the former (also called syncytin-1 and syncytin-2 and likely to be involved in human placentation) and provide tools for the latter, to determine their potential role in physiological processes and/or their association with pathological processes in humans – which would be the consequence of their original retroviral status.
Methods
DNA samples and genotyping
Ninety-one human samples of French Caucasians were collected from the EGEA (Epidemiological study on the Genetics and Environment of Asthma) study, among the controls ascertained without disease. PCR was performed in mixture containing 25 ng of DNA, 0.3 pmol of each primer, 6 nmol of each dNTP, 0.75 units of ExTaq and 1× reaction buffer (Takara). Sequencing reactions were performed according to the Dye Terminator method using an ABI PRISM® 3700 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). Alignment of sequences, SNP discovery and genotyping were performed with Genalys software [29]. The genomic sequences used for the alignment are envFRD (GenBank accession no. AL136139), envR (AC073210), envT (AC078899), envW (AC000064), envFc(1) (AL354685), envFc(2) (AC01222), envH1 (AJ289709), envH2 (AJ289710), envH3 (AJ289711), envR(b) (AC093488), envK1 (AC074261), envK2 (AC072054), envK4 (AF164615). The sources of the simian genomic DNAs are given in ref [24].
Statistical analysis
The haplotypes frequencies using all polymorphisms for each gene were estimated with the EM Algorithm [20]. The linkage disequilibrium (LD) estimates between pairs of polymorphisms were obtained by estimating the two polymorphisms haplotypes frequencies using this algorithm. Computation of D and D' (standard disequilibrium measure and standardized disequilibrium measure) in additional data file 3 was performed as in ref [30].
Cloning of allelic forms of the W and FRD human env genes in expression vectors
The FRD and W env genes were PCR-amplified from human genomic DNAs. PCR was carried out for 25 cycles (10 sec at 93°C, 30 sec at 56°C, 4 min at 68°C), in 50 μl, using 100 ng of genomic DNA, 48 pmol of each primer, 350 μM of each dNTP, 0.75 μl Expand long template enzyme mix and 1× reaction buffer (Roche Applied Science). For the FRD env gene, XhoI-containing primers were ATCACCTCGAGCACCATGGGCCTGCTCCTGCTGGTTCTCATTC as forward primer and ATCACCTCGAGGCTTCAGTACAGGTGGATA as reverse primer. For the W env gene, XhoI-containing primers were ATCACCTCGAGAACAACCAGGAGGAAAGTAA as forward primer and ATCACCTCGAGCTGATCAAGTCGCAAAGC as reverse primer. Each PCR product was then XhoI-cleaved and cloned into the phCMV-G vector (described in [12]) opened with XhoI. Allelic forms of the cloned env genes were assessed by enzymatic restriction. For envFRD, the 1075 G->A transition (aa 359) predicts the loss of a RsaI site and the 1100 C->T transition (aa 367) predicts the loss of a BstUI site, and for envW the 413 G->A transition (aa 138) predicts the gain of a BstXI site and the 920 G->A transition (aa 307) predicts the gain of a Tsp509I site. As two allelic forms (FRD359T/367M and W138Q/307S) were not available among the cloned envelope genes, we constructed them by exchange of restriction fragments (BsmI-NotI for FRD359T/367M and KpnI-XhoI for W138Q/307S).
Cell-cell fusion assay
The human TE671 rhabdomyosarcoma cells (ATCC CRL8805) and 293T embryonal kidney cells (ATCC CRL11268) were grown in Dulbeco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum. All cell culture media were supplemented with streptomycin (100 μg/mL) and penicillin (100 U/mL). Cells were transfected using calcium phosphate precipitation (Invitrogen, 5 μg of DNA for 5 × 105 cells). Fusion activity of envelope glycoproteins was measured 12 to 36 h after transfection of the cells with the corresponding expression vectors. To visualize syncytia, cells were fixed in methanol and stained by adding May-Grünwald and Giemsa solutions (Sigma) according to the manufacturer's instructions. The fusion index, which represents the percentage of fusion events in a cell population is defined as [(N-S)/T] × 100, where N is the number of nuclei in the syncytia, S is the number of syncytia, and T is the total number of nuclei counted.
Characterization of the orthologous envT, envF(c)2 and envR(b) env gene ORFs from simians
The size of the env gene open reading frame in the primate loci was evaluated using a direct coupled in vitro transcription/translation assay based on T7 promoter-containing PCR products as described in [24], which allows to determine the status of both alleles in the same assay.
For the amplification of envT from gorilla and orangutan, the forward T7 promoter-containing primers were GCTAATACGACTCACTATAGGAACAGACCACCATGTCCTGCTTGGATTCATCAC and GCTAATACGACTCACTATAGGAACAGACCACCATGTTGGATTCATCACTCCCA, respectively, and the common reverse flanking primer was CTGAAGGGAGTTCCTCCTAGG. For the amplification of envR(b) from chimpanzee, gorilla, orangutan, gibbon, Rhesus macaque (Old World Monkey) and Callithrix jacchus (New World Monkey) the forward T7 promoter-containing primer was GCTAATACGACTCACTATAGGAACAGACCACCATGGATCCACTACACACGATTGA and the reverse flanking primer was TGTTTTGGGACACCACGAAT. For the amplification of envF(c)2 from chimpanzee and gorilla, the forward T7 promoter-containing primer was GCTAATACGACTCACTATAGGAACAGACCACCATGAATTCTCCATGTGAC and the reverse flanking primer was GACACTTAATAGTTGCGACA.
The simian env gene sequences were deposited in Genbank with accession numbers AJ862646-AJ862655.
List of abbreviations used
HERV, human endogenous retrovirus; TM, transmembrane; SNP, single nucleotide polymorphism; LTR, Long Terminal Repeat; LD, linkage disequilibrium.
Authors' contributions
NdP designed the cloning primers for the env genes, carried out the interspecific sequence conservation studies and drafted the manuscript.
GD carried out the SNP studies.
SB carried out the cloning of the env genes and the cell-cell fusion assays.
FH participated in the interspecific sequence conservation studies.
AV designed the sequencing primers, participated in the coordination of the SNP studies and in drafting the manuscript.
FM coordinated the SNP studies.
TH conceived the study.
Supplementary Material
Additional data file 1
Additional data file 1 is a table listing the polymorphism of HERV coding envelope genes.
Click here for file
Additional data file 2
Additional data file 2 is a figure showing polymorphism combinations and estimated frequencies for haplotypes in the 12 HERV coding env genes. The SNPs are identified by their CNG ID (see additional data file 1).
Click here for file
Additional data file 3
Additional data file 3 is a figure showing the LD plot of 12 HERV coding env genes. The LD pattern is shown with the D values above and the D' values below the diagonal, and estimated allele frequencies for each polymorphism. Different colors are used to represent ranges of positive D and D' values. The SNPs are identified by their CNG ID (see additional data file 1).
Click here for file
Acknowledgements
This work was supported by the CNRS and by grants from the Ligue Nationale contre le Cancer (Equipe Labellisée for T.H.). We thank Evelyne Heyer for helpful discussions and acknowledge Christian Lavialle for critical reading of the manuscript.
==== Refs
Consortium IHGS Initial sequencing and analysis of the human genome Nature 2001 409 860 921 11237011 10.1038/35057062
Boeke JD Stoye JP Coffin JM, Hughes SH and Varmus HE Retrotransposons, endogenous retroviruses, and the evolution of retroelements Retroviruses 1997 New York, Cold Spring Harbor Laboratory Press 343 436
Bannert N Kurth R Retroelements and the human genome: New perspectives on an old relation Proceedings of the National Academy of Sciences of the United States of America 2004 13 Suppl 2 14572 14579 15310846 10.1073/pnas.0404838101
de Parseval N Heidmann T Human endogenous retroviruses: from infectious elements to human genes Cytogenet Genome Res 2005 110 318 332 16093684 10.1159/000084964
Repbase update, a database of transposable elements
de Parseval N Lazar V Casella JF Benit L Heidmann T Survey of human genes of retroviral origin: identification and transcriptome of the genes with coding capacity for complete envelope proteins Journal of Virology 2003 77 10414 10422 12970426 10.1128/JVI.77.19.10414-10422.2003
Best S Le Tissier PR Stoye JP Endogenous retroviruses and the evolution of resistance to retroviral infection Trends in Microbiology 1997 5 313 318 9263409 10.1016/S0966-842X(97)01086-X
Cianciolo GJ Copeland T Orozlan S Snyderman R Inhibition of lymphocyte proliferation by a synthetic peptide homologous to retroviral envelope protein Science 1985 230 453 455 2996136
Mangeney M Heidmann T Tumor cells expressing a retroviral envelope escape immune rejection in vivo Proc Natl Acad Sci USA 1998 95 14920 14925 9843991 10.1073/pnas.95.25.14920
Blond JL Lavillette D Cheynet V Bouton O Oriol G Chapel-Fernandes S Mandrand B Mallet F Cosset FL An envelope glycoprotein of the human endogenous retrovirus HERV-W is expressed in the human placenta and fuses cells expressing the type D mammalian retrovirus receptor Journal of Virology 2000 74 3321 3329 10708449 10.1128/JVI.74.7.3321-3329.2000
Mi S Lee X Li X Veldman GM Finnerty H Racie L LaVallie E Tang XY Edouard P Howes S Keith JCJ McCoy JM Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis Nature 2000 403 785 788 10693809 10.1038/35001608
Blaise S de Parseval N Bénit L Heidmann T Genomewide screening for fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene conserved on primate evolution Proceedings of the National Academy of Sciences of the United States of America 2003 100 13013 13018 14557543 10.1073/pnas.2132646100
Frendo JL Olivier D Cheynet V Blond JL Bouton O Vidaud M Rabreau M Evain-Brion D Mallet F Direct involvement of HERV-W Env glycoprotein in human trophoblast cell fusion and differentiation Mol Cell Biol 2003 23 3566 3574 12724415 10.1128/MCB.23.10.3566-3574.2003
de Parseval N Heidmann T Physiological knock-out of the envelope gene of the single copy ERV-3 human endogenous retrovirus in a fraction of the caucasian population Journal of Virology 1998 72 3442 3445 9525678
Turner G Barbulescu M Su M Jensen-Seaman MI Kidd KK Lenz J Insertional polymorphisms of full-length endogenous retroviruses in humans Current Biology 2001 11 1531 1535 11591322 10.1016/S0960-9822(01)00455-9
Hughes JF Coffin JM Human endogenous retrovirus K solo-LTR formation and insertional polymorphisms: implications for human and viral evolution Proc Natl Acad Sci U S A 2004 101 1668 1672 14757818 10.1073/pnas.0307885100
Zhao Z Fu YX Hewett-Emmett D Boerwinkle E Investigating single nucleotide polymorphism (SNP) density in the human genome and its implications for molecular evolution Gene 2003 312 207 213 12909357 10.1016/S0378-1119(03)00670-X
Stephens JC Schneider JA Tanguay DA Choi J Acharya T Stanley SE Jiang R Messer CJ Chew A Han JH Duan J Carr JL Lee MS Koshy B Kumar AM Zhang G Newell WR Windemuth A Xu C Kalbfleisch TS Shaner SL Arnold K Schulz V Drysdale CM Nandabalan K Judson RS Ruano G Vovis GF Haplotype variation and linkage disequilibrium in 313 human genes Science 2001 293 489 93. Epub 2001 Jul 12. 11452081 10.1126/science.1059431
Mallet F Bouton O Prudhomme S Cheynet V Oriol G Bonnaud B Lucotte G Duret L Mandrand B The endogenous retroviral locus ERVWE1 is a bona fide gene involved in hominoid placental physiology Proc Natl Acad Sci U S A 2004 101 1731 1736 14757826 10.1073/pnas.0305763101
Laird N Rao CR Computational Statistic: the EM algorithm Handbook of Statistics 1993 9 Amsterdam, Elsevier Science Publishers 509–520 10.1016/S0169-7161(05)80138-5
Blaise S Ruggieri A Dewannieux M Cosset FL Heidmann T Identification of an envelope protein from the FRD family of Human Endogenous Retroviruses (HERV-FRD) conferring infectivity on retroviral particles and functional conservation among simians Journal of Virology 2004 78 1050 1054 14694139 10.1128/JVI.78.2.1050-1054.2004
Herve CA Forrest G Lower R Griffiths DJ Venables PJ Conservation and loss of the ERV3 open reading frame in primates Genomics 2004 83 940 943 15081124 10.1016/j.ygeno.2003.10.003
Benit L Calteau A Heidmann T Characterization of the low-copy HERV-Fc family: evidence for recent integrations in primates of elements with coding envelope genes Virology 2003 312 159 168 12890629 10.1016/S0042-6822(03)00163-6
de Parseval N Casella JF Gressin L Heidmann T Characterization of the three HERV-H proviruses with an open envelope reading frame encompassing the immunosuppressive domain and evolutionary history in primates Virology 2001 279 558 569 11162811 10.1006/viro.2000.0737
Fay JC Wu CI Sequence divergence, functional constraint, and selection in protein evolution Annu Rev Genomics Hum Genet 2003 4 213 235 14527302 10.1146/annurev.genom.4.020303.162528
Bonnaud B Bouton O Oriol G Cheynet V Duret L Mallet F Evidence of selection on the domesticated ERVWE1 env retroviral element involved in placentation Mol Biol Evol 2004 21 1895 1901 15254254 10.1093/molbev/msh206
Zhang J Webb DM Evolutionary deterioration of the vomeronasal pheromone transduction pathway in catarrhine primates Proc Natl Acad Sci U S A 2003 100 8337 8341 12826614 10.1073/pnas.1331721100
Belshaw R Pereira V Katzourakis A Talbot G Paces J Burt A Tristem M Long-term reinfection of the human genome by endogenous retroviruses Proc Natl Acad Sci U S A 2004 101 4894 4899 15044706 10.1073/pnas.0307800101
Takahashi M Matsuda F Margetic N Lathrop M Automated identification of single nucleotide polymorphisms from sequencing data. J Bioinform Comput Biol 2003 1 253 265 15290772 10.1142/S021972000300006X
Devlin B Risch N A comparison of linkage disequilibrium measures for fine-scale mapping Genomics 1995 29 311 322 8666377 10.1006/geno.1995.9003
|
16150157
|
PMC1236922
|
CC BY
|
2021-01-04 16:32:48
|
no
|
BMC Genomics. 2005 Sep 9; 6:117
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-117
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-571613139310.1186/1472-6963-5-57Research ArticleHeart failure guidelines and prescribing in primary care across Europe Sturm Heidrun B [email protected] Gilst Wiek H [email protected] Karl [email protected] FD Richard [email protected] Flora M [email protected] Department of Clinical Pharmacology, University Medical Center Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands2 Göteborg University, Department of Medicine, Sahlgrenska University Hospital, SE-416 85 Göteborg, Sweden3 Department of Primary Care and General Practice, Medical School, University of Birmingham, Birmingham, B15 2TT UK2005 30 8 2005 5 57 57 29 11 2004 30 8 2005 Copyright © 2005 Sturm et al; licensee BioMed Central Ltd.2005Sturm 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
Major international differences in heart failure treatment have been repeatedly described, but the reasons for these differences remain unclear. National guideline recommendations might be a relevant factor. This study, therefore, explored variation of heart failure guideline recommendations in Europe.
Methods
Treatment recommendations of 14 national guidelines published after 1994 were analyzed in relation to the heart failure treatment guideline of the European Society of Cardiology. To test potential relations between recommendations and prescribing, national prescribing patterns as obtained by a European study in primary care (IMPROVEMENT-HF) were related to selected recommendations in those countries.
Results
Besides the 14 national guidelines used by primary care physicians in the countries contacted, the European guideline was used in four countries, and separate guidelines for specialists and primary care were available in another four countries. Two countries indicated that no guideline was used up to 2000. Comprehensiveness of the guidelines varied with respect to length, literature included and evidence ratings. Relevant differences in treatment recommendations were seen only in drug classes where evidence had changed recently (β-blockers and spironolactone). The relation between recommendation and prescribing for selected recommendations was inconsistent among countries.
Conclusion
Differences in guideline recommendations are not sufficient to explain variation of prescribing among countries, thus other factors must be considered.
==== Body
Background
Chronic heart failure is a common disease in all developed countries, and its prevalence is likely to increase further due to aging societies and improvement in therapies [1]. Therefore, the interest in good quality of heart failure management is twofold: optimal medical outcomes as well as the efficient use of resources.
Despite internationally available identical evidence, differences in heart failure therapy among countries have been repeatedly described for inpatient as well as outpatient care [2-4]. In the European Improvement of Heart Failure survey (IMPROVEMENT-HF) for instance, β-blocker-use in primary care ranged from 10% in Turkey, to 26% in the UK and 50% or more in Sweden and Hungary. ACE-use ranged from almost 50% to 75%. Since available evidence is translated into national guidelines in most countries, national recommendations may differ and accordingly contribute to such variation.
Differences among guidelines have been shown for various diseases. While actual content varied little, the major variation was found in the method and rigor of guideline development as well as the comprehensiveness [5-8]. The focus of this study is on content of published guidelines in Europe. Firstly we assess the variation in recommendations for heart failure treatment; secondly we explore their association with prescribing patterns in European countries as observed in the IMPROVEMENT-HF survey.
Methods
Only guidelines from European countries (Belgium, Czech Republic, France, Germany, Great Britain, Hungary, Italy, Portugal, Russia, Scotland, Spain, Sweden, Switzerland, The Netherlands, Turkey) were considered. Guidelines were compared to the treatment guideline of the European Society of Cardiology (ESC) [9]. The version of 1997 was chosen as the reference guideline in order to relate recommendations to prescribing data from the European IMPROVEMENT-HF survey, which was conducted in 1999/2000 [3]. This study reported prescribing data from 11062 patients treated by 2327 primary care physicians in 14 countries.
Guidelines and recommendations for heart failure were collected through national experts (members of the ESC working group on heart failure or a recommended substitute) and by search of websites of national cardiology societies as well as Medline. Only guidelines produced by national organizations or published on a national level between 1994 and 2002 were included. Regional guidelines were not taken into account. For the main analysis the original version was always used, even if updates became available.
Two approaches were used to determine the content of national guidelines: Firstly, national experts completed a questionnaire about the content of the most commonly used guideline by primary care physicians (PCPs) in their country at the time of the IMPROVEMENT-HF study (up to 2000). Secondly, the researcher together with a native speaker with medical background used the original guideline documents to complete the questionnaire. In case of discordance, the national expert was contacted again.
As a proxy for quality, four dimensions of formal appraisal instruments [10,6] were used (provided information about authors, specified target group, information about used evidence (citations), grading of evidence).
Form and content of guidelines with different publication years or target groups were compared. Information about publication, dissemination and use of all existing national heart failure treatment guideline(s) was retrieved from the questionnaire and described.
Obtained recommendations were compared to the ESC recommendations and grouped into one of the following three categories: recommendations identical with the ESC recommendation (identical); recommendations differing from ESC (disagreement); and ESC recommendations not included or mentioned in the particular guideline (not specified). In case of disagreement specification was asked.
To relate recommendations to prescribing only recommendations of guidelines available before the survey were used. If more than one guideline was published in a country, the one mostly used by GPs (Italy, Germany and NL) was included. Prescribing data from the IMPROVEMENT-HF survey was used. Prescribing between groups of countries with identical recommendations was compared and the effect of recommendations on prescribing was analyzed using logistical regression (SPSS 11.0). Multivariate analysis was performed to account for patient characteristics influencing prescribing (age, sex and severity of disease (NYHA); significant at the 5% level in univariate analysis).
Results
From 16 heart failure experts contacted, 13 replied citing 16 guidelines. Besides the European reference guideline, a further 14 guidelines for treatment of heart failure were included in the study (Table 1). The Scottish and the Russian documents were retrieved from the Internet and no expert opinion was available. Belgium and Turkey reported that no CHF-guideline had been used in their country. In four countries (Hungary, Portugal, Switzerland, UK) the ESC guideline was used exclusively. More than one guideline had been published in four countries, different for specialists and primary care. According to expert information, guidelines were disseminated via publication in professional literature (14 of 16), by scientific meetings (14 of 16) and Continuing Medical Education (CME; 11 of 16). Three guidelines have been updated along with the changes to the ESC guideline.
Table 1 List and characteristics of analyzed guidelines for heart failure
Country Pub year Authors Dev. Target group No. Ref Grading Epidem Etiology Econ. relevance
NL 1994 CBO + Specialists 0 - + + +
NL 1995 NHG + GP 191 + + + -
France 1996 Individual experts - Specialists 129 + + + +
Hungary Portugal CH, UK 1997 Task Force of European Society of Cardiology + All physicians* 69 + + + +
Italy 1997 Task force in collaboration with society of non-hospital-cardiologists, SIC + Specialists 120 + + + -
CZ 1998 Czech Society of Cardiology GP 30 - + + -
D 1998 Medical Commission of the physician associations (AkdA) + GP 66 + + + -
D 1998 German Society of Cardiology, DGK - Specialists 213 + + + -
Italy 1998 ANMCO (Cardiol. Society) and SIMG (GP-society) - GP 0 - - - -
Scotland 1998 SIGN (Scottish Intercollegiate Guideline Network) + Not specified 211 + + + +
Sweden 1998 SOS (National Board of Health and Welfare) + All physicians 0 - + + -
Spain 1999 Working Group for HF of the Spanish Soc. of Cardiology + GP and specialists 113 + + + +
Sweden 2000 MPA (Medical Product Agency) Physicians other than cardiologists 0 + + + +
Russia 2001 Ju. N. Belenkow, Scientific Research Institute of Cardiology, Ministry of Health, Moscou - Not specified 0 - + + +
CH 2002 Working group on HF of the Swiss society of cardiology + Not specified 24 + + + +
Total (n = 15) 9 9 10 10 13 13 8
+ included, - not included, GP = General Practitioner, NL: The Netherlands, CZ: Czech Republic, CH: Switzerland, D: Germany, Swe: Sweden
Dev.: method of development is outlined in the document. Target group: as specified in the GL, according to publication or as indicated by specialist in questionnaire.
Grading: levels of scientific evidence given with recommendations. "Quality" included specification of development, target group, evidence levels and literature.
* In Portugal the ESC-guideline was targeted at specialists. # This version was last updated 11/2001
Form
Presentation and comprehensiveness of the guidelines varied (Table 1). Length along with included literature, ranged from comprehensive publications to handouts comprising only flowcharts. The method of development was clearly stated in 8, evidence ratings were included in 10 guidelines. National specialist societies or physician organizations were the authors in all but two cases. Epidemiology was discussed in all guidelines, etiology in all but the Italian leaflet for general practitioners. The economic relevance of heart failure was mentioned in about half of the documents. In 10 cases a target group of physicians was specified in the guidelines or was defined by the publishing organization or journal, which addressed specific groups of physicians. All guidelines specifically directed at specialists (except the oldest, the Dutch CBO) included more than 100 citations and gave evidence levels for their recommendations, whereas, on the other hand, guidelines without evidence levels or less than 30 citations were directed at PCPs or did not specify the target group (Table 1). Guidelines for specialists contained more detailed recommendations than the others (10% non-specified recommendations in guidelines for specialists as opposed to 24% in the others, Table 2), yet the content of included recommendation did not differ vastly. Overall, three of the 15 guidelines provided information on all four quality dimensions, six did not specify either the target group or the development procedure but all the remaining criteria.
Table 2 Overall differences in content between guidelines for different target groups
recommendations* Identical with ESC [%] Disagreement [%] Not specified [%]
Guidelines for† Specialists (n = 5) 70 19 10
other (n = 9) 63 13 24
All (n = 14) 65 16 19
*Each guideline was checked for 16 recommendations and compared to the European guideline (ESC) of 1997
† see Figure 1.
Content
Similarities
Drug therapy recommendations for diuretics, ACE-inhibitors and glycosides did not reveal major differences (Figure 1). Only the Dutch NHG-guideline for primary care recommended diuretics as first drug choice prior to ACE inhibitors. The cut-off value for the ejection fraction below which ACE-treatment should be initiated varied between 40% (ESC) to 25% (Russia).
Figure 1 Variation in recommendations included in European guidelines. Recommendations relating to one drug are grouped; corresponding results (percentages of identical, different or not specified recommendations) are shown in adjacent horizontal bars. n = 14 national guidelines (other than the European Society of Cardiology (ESC)-guideline).
Differences
The indications for β-blocker use varied (Figure 1). Whereas 9 guidelines recommended their use for all stable CHF patients (three excluded most severe cases, NYHA class IV), the ESC guideline along with three others advocated β-blockers only for idiopathic dilated cardiomyopathy. Those were published in 1994, 1998 and 2000.
The recommendation to restrict their initiation and control to specialists was included by the majority, with an overall range from: not to be used by GPs at all (Dutch NHG 1994) to: by any doctor (Spain, 2000). The relation to publication time was generally consistent: The limitation to specialists was included mainly in earlier guidelines and the two guidelines clearly disagreeing were both from 2000. Yet, in Switzerland the limitation was still included in 2002.
Also recommendations for spironolactone showed differences (Figure 1): The Dutch CBO (1994) and the Italian brochure (1997) did not mention spironolactone at all. Before 1999 agreement about the use of spironolactone as an additional diuretic in hypokalemia was more frequent (5 of 10 vs. 1 of 4 after 1999). Also more frequent was the recommendation not to combine spironolactone with ACE. After 1999 this warning was replaced by the recommendation to use spironolactone in addition to ACE and other therapy in severe heart failure.
Only minor variation was found in recommendations for other drug groups: The combination of nitrates and hydralazine as additional symptomatic treatment choice was only recommended by the two German guidelines. Of the 9 guidelines addressing anticoagulation, three specified their indication for patients with an EF < 20–25% (D, Italy) or NYHA 3 and 4 (Russia).
Updates in guidelines
When comparing the different versions of the German, the Dutch CBO and the ESC guideline, the emergence of new evidence for β-blockers and spironolactone is reflected by the order in which it is discussed in the guidelines: β-blockers in the old versions were the fourth or later topic, in the new versions they were moved to second or third. Spironolactone was no longer discussed within the diuretics-section but separately. For other drug groups no major changes were found. In the Dutch document the use of tests and particularly echocardiography to establish diagnosis was more clearly recommended.
Guidelines and prescribing
Recommendations for ACE-inhibitors did not differ between the countries, yet prescribing ranged from 48% in Sweden and 76% in Hungary[3]. Also glycoside-use ranged from 24% in the UK to 45% in Italy and 55% in Turkey.
Where major differences in recommendations were present (for spironolactone and β-blockers), the relation to prescribing was inconsistent. Where recommendations warned about the combination of spironolactone with ACE (before 2000, average year of publication: 1997) this drug-combination was prescribed significantly less than where no warning was included in the guideline (Table 3).
Table 3 Actual prescribing in relation to recommendations
Recommendation Countries
(% of total) Year of publication
Range (mean) (%) Drug use
% patients with OR (95%CI)
Spironolactone in general not to be given in combination with ACE ACE + spironolactone
univariate multivariate
All countries* 10381
(100%) 6.6
Warning (Identical) NL, Cz, F, D, Swe
(36,5%) 1995–1998
(1997) 5,7 0.78
(0.66–0.91) 0.74
(0.63–.87)
no warning/no recommendation† Remaining
(63,5%) 1998
(1998) 7,2 1 1
β-blockers are indicated only for patients with IDCM (idiopathic dilated cardiomyopathy) β-blocker
All countries* 10381
(100%) 33.8
Restricted indication (identical) Cz, H, CH, UK
(28.6%) 1998
(1998) 40.4 1 1
No restriction Remaining†
(71.4%) 1995–1998
(1997) 31.1 0.67
(0.61–0.73) 0.69
(0.63–0.76)
*Poland was excluded from the analysis, because no GL information was available.
†includes all countries without guideline before 2000 and guidelines without recommendation. In the case of NL and Italy, the specific guideline for GPs has been used.
In contrast, β-blockers were prescribed more likely if recommendations restricted their indication to the subgroup of patients with idiopathic dilated cardiomyopathy (Table 3). Those guidelines (available before 2000) were all published 1998 with the exception of the Dutch guidelines for specialists from 1994; the ones without specification were published between 1995 and 1998. Also if the Dutch and Italian guidelines for specialists were used in the statistical model, the same trend was found. In both cases correcting for patient characteristics did not change the pattern either.
Discussion
The goal of this study was to assess the degree of variation between recommendations for heart failure treatment in Europe as an potential factor explaining international variation in prescribing. To our knowledge the content of a wide range of European heart failure guidelines and its relation to actual prescribing data has not been assessed before [5].
Similar to other studies comparing guideline recommendations [8,6] our results did not reveal major differences. With the exception of one of the oldest, all guidelines recommended ACE-inhibitors as first line medication. Recommendations for diuretics and glycosides were basically identical.
Most variance was found where the emergence of new evidence induced changes in established therapy during the study period, namely for spironolactone [11] and β-blockers [12]. The role of spironolactone in severe heart failure was changed by the RALES study in 1999. This is reflected in a switch from warning about the combination of spironolactone with ACE inhibitors in earlier guidelines to recommending this combination in the later ones. The growing importance of β-blockers for CHF therapy was reflected in a trend to widen the indication and not to restrict its prescribing to specialists any more. However, the respective conservative recommendation was still included in one newer guideline (from 2000 or 2002). On the other hand different recommendations were not reflected in the average publication year of guidelines. Thus, new evidence is not always taken up at the same pace; some countries appear to be more conservative or cautious than others.
Still, as to be expected the frequency of updates influenced the time lag between emergence and inclusion of new evidence. The two German guidelines were updated within three years and new evidence was reflected in the changes made. In the Netherlands the guideline for primary care from 1995 was updated together with the 1994 guideline for specialists in an interdisciplinary manner only in 2002. Thus, despite being early in guideline development, this time span between updates resulted in some recommendations, which were not in accordance with newer evidence for some years.
The target group to which a guideline is directed appeared to play a role for both, the formal scientific appearance and the content. Guidelines for primary care physicians included fewer citations, gave evidence levels less frequently, and included fewer specific recommendations. It has been shown, that form and credibility of the guidelines' authors influence their uptake [13,14]. Given the fact, that primary care is where prescribing for heart failure mainly takes place in most European countries [15,16] and that primary care physicians tend to be less in accordance with newer evidence than specialists [17-20], guidelines directed at general practitioners should be given special attention. In our study, in six countries no specific guideline for primary care physicians could be detected. Thus, to adjust form and content of guidelines to primary care physicians' needs might be one way of improving the implementation of evidence in primary care.
By associating national recommendations with prescribing data we aimed to further explore the role of guidelines in international variation of heart failure therapy. The found relationship, however, appeared to be inconsistent: On the one hand, guideline-recommendations for major drug classes were similar while prescribing e.g. for ACE-inhibitors differed by about 25% among countries. On the other hand, when recommendations differed between countries, prescribing was not always correlated. Another example from our data for the weak link between prescribing and guideline recommendations is the Dutch case, where despite the conservative guideline recommendations in 1999 the frequency of ACE and β-blocker prescribing ranged at the European mean.
Limitations to the study
In order to better understand the relationship between heart failure prescribing and national recommendations in European primary care, this study focused on published content rather than on the rigor of development of guidelines. In view of this focus guidelines were assessed in a explorative, descriptive manner. Instead of using one formal guideline appraisal instrument [21] we assessed major dimensions according to the checklists provided by Grilli [10] and Kulig [6] to give an indication about quality. In depth analysis of recommendations was limited to major drug groups and therefore minor differences might not have been considered.
Only European guidelines were included in this study, as they were to be analyzed in connection with the prescribing data. For the same reason the European guideline of 1997 (ESC) formed the basis for the content analysis rather than the later version of 2001. Nevertheless newer recommendations were included and therefore changes over time could still be detected.
In the statistical analysis, we accounted for differences in selected patient characteristics between the compared groups that are relevant determinants of prescribing, however other potentially relevant patient (or doctor) factors were not included. In addition, in this cross sectional analysis differences in time of guideline publication and emergence of new evidence could not be accounted for. Therefore prescribing data and its link to recommendation can only give an indication about general relations between the two.
Further, countries can only be grouped according to published national guidelines. This might not in all cases reflect recommendations actually used by physicians.
Detailed information about the implementation process [22-24] as a major aspect for a guideline's impact on prescribing was beyond the scope of this study, although our expert replies suggest roughly comparable approaches of guideline dissemination. The range of different forms as well as of authoring groups may have further compromised comparability of acceptance and uptake.
Although recruitment procedures of practices aimed to include comparable national populations, variations in physician characteristics were still present. Further, even though only primary care practices were included, there might have been relevant differences in national organization of care.
Conclusion
In the majority of the European countries surveyed national heart failure treatment guidelines were available at the time of the study. Nine of 15 guidelines included information about at least three quality dimensions (specified authors and development, specified target group, citations and grading of evidence), still form and comprehensiveness varied substantially.
Differing degrees of comprehensiveness for different user groups suggest that increased attention to the target group of primary care when issuing recommendations might enhance the uptake of guidelines, bearing in mind that CHF is treated primarily in primary care.
This study suggests that national guidelines only play a minor role in explaining differences in heart failure therapy between countries. The majority of recommendations were similar, differences were found mainly within drug groups where evidence has undergone recent change. Also, in the case of distinct national recommendations, prescribing within countries does not appear to consistently follow the advice. These data emphasize the relevance of country specific factors for prescribing at a national level, one of which may be the implementation of guideline recommendations.
Competing interests
HS, WvG, FMH-R, KS declare that they have no competing interests.
FDRH has received fees, research funding or sponsorship, on an occasional basis, from a number of pharmaceutical or diagnostics companies active in the field of heart failure.
Authors' contributions
HS, FMH, WG participated in the study design. HS and FMH developed the questionnaire; HS collected the guidelines, followed up with experts, analyzed the documents with native speakers and drafted the manuscript. FRH, KS, WG and FMH advised which experts to contact, helped with interpretation of results and with the draft. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
Individual recommendations per country. part a-c gives country wise results for each recommendation.
Click here for file
==== Refs
Cleland JG Khand A Clark A The heart failure epidemic: exactly how big is it? Eur Heart J 2001 22 623 626 11286518 10.1053/euhj.2000.2493
Komajda M Follath F Swedberg K Cleland J Aguilar JC Cohen-Solal A Dietz R Gavazzi A van Gilst WH Hobbs R Korewicki J Madeira HC Moiseyev VS Preda I Widimsky J Freemantle N Eastaugh J Mason J The EuroHeart Failure Survey programme--a survey on the quality of care among patients with heart failure in Europe. Part 2: treatment Eur Heart J 2003 24 464 474 12633547 10.1016/S0195-668X(02)00700-5
Cleland JG Cohen-Solal A Aguilar JC Dietz R Eastaugh J Follath F Freemantle N Gavazzi A van Gilst WH Hobbs FD Korewicki J Madeira HC Preda I Swedberg K Widimsky J Management of heart failure in primary care (the IMPROVEMENT of Heart Failure Programme): an international survey Lancet 2002 360 1631 1639 12457785 10.1016/S0140-6736(02)11601-1
Bellotti P Badano LP Acquarone N Griffo R Lo PG Maggioni AP Mattiauda C Menardo G Mombelloni P Specialty-related differences in the epidemiology, clinical profile, management and outcome of patients hospitalized for heart failure; the OSCUR study. Oucome dello Scompenso Cardiaco in relazione all'Utilizzo delle Risore Eur Heart J 2001 22 596 604 11259147 10.1053/euhj.2000.2362
Grimshaw GM Khunti K Baker R Diagnosis of heart failure in primary care: an assessment of international guidelines Br J Gen Pract 2001 51 384 386 11360704
Kulig M Erika S Norbert WS Comparing methodological quality and consistency of international guidelines for the management of patients with chronic heart failure Eur J Heart Fail 2003 5 327 335 12798831 10.1016/S1388-9842(03)00040-0
Burgers JS Bailey JV Klazinga NS Van Der Bij AK Grol R Feder G Inside guidelines: comparative analysis of recommendations and evidence in diabetes guidelines from 13 countries Diabetes Care 2002 25 1933 1939 12401735
Iqbal A Schloss S George D Isonaka S Worldwide guidelines for chronic obstructive pulmonary disease: A comparison of diagnosis and treatment recommendations Respirology 2002 7 233 239 12153689 10.1046/j.1440-1843.2002.00399.x
The treatment of heart failure. The Task Force on Heart Failure of the European Society of Cardiology Eur Heart J 1997 18 736 753 9152644
Grilli R Magrini N Penna A Mura G Liberati A Practice guidelines developed by specialty societies: the need for a critical appraisal Lancet 2000 355 103 106 10675167 10.1016/S0140-6736(99)02171-6
Pitt B Zannad F Remme WJ Cody R Castaigne A Perez A Palensky J Wittes J The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators N Engl J Med 1999 341 709 717 10471456 10.1056/NEJM199909023411001
A randomized trial of beta-blockade in heart failure. The Cardiac Insufficiency Bisoprolol Study (CIBIS). CIBIS Investigators and Committees Circulation 1994 90 1765 1773 7923660
Burgers JS Grol RPTM Zaat JOM Spies TH Van Der Bij AK Mokkink HGA Characteristics of effective clinical guidelines for general practice British Journal of General Practice 2003 53 15 19 12569898
Remme WJ Towards the better treatment of heart failure Eur Heart J 1998 19 Suppl L L36 L42 9821008
Cline CM Boman K Holst M Erhardt LR Failure fS SS The management of heart failure in Sweden Eur J Heart Fail 2002 4 373 376 12034164
Komajda M Bouhour JB Amouyel P Delahaye F Vicaut E Croce I Rougemond E Vuittenez F Leutenegger E Ambulatory heart failure management in private practice in France Eur J Heart Fail 2001 3 503 507 11511438 10.1016/S1388-9842(01)00172-6
McKee SP Leslie SJ LeMaitre JP Webb DJ Denvir MA Management of chronic heart failure due to systolic left ventricular dysfunction by cardiologist and non-cardiologist physicians Eur J Heart Fail 2003 5 549 555 12921818 10.1016/S1388-9842(03)00003-5
Rutten FH Grobbee DE Hoes AW Differences between general practitioners and cardiologists in diagnosis and management of heart failure: a survey in every-day practice Eur J Heart Fail 2003 5 337 344 12798832 10.1016/S1388-9842(03)00050-3
Fuat A Hungin AP Murphy JJ Barriers to accurate diagnosis and effective management of heart failure in primary care: qualitative study BMJ 2003 326 196 12543836 10.1136/bmj.326.7382.196
Khunti K Hearnshaw H Baker R Grimshaw G Heart failure in primary care: qualitative study of current management and perceived obstacles to evidence-based diagnosis and management by general practitioners Eur J Heart Fail 2002 4 771 777 12453549 10.1016/S1388-9842(02)00119-8
Cluzeau F Burgers J Brouwers M Grol R Makela M Littlejohns P Grimshaw J Hunt C Development and validation of an international appraisal instrument for assessing the quality of clinical practice guidelines: the AGREE project Quality & Safety in Health Care 2003 12 18 23 12571340 10.1136/qhc.12.1.18
Adams KF Translating heart failure guidelines into clinical practice: clinical science and the art of medicine Curr Cardiol Rep 2001 3 130 135 11177670
Wensing M van der WT Grol R Implementing guidelines and innovations in general practice: which interventions are effective? Br J Gen Pract 1998 48 991 997 9624774
van Veldhuisen DJ Charlesworth A Crijns HJ Lie KI Hampton JR Differences in drug treatment of chronic heart failure between European countries Eur Heart J 1999 20 666 672 10208787 10.1053/euhj.1998.1343
|
16131393
|
PMC1236923
|
CC BY
|
2021-01-04 16:31:51
|
no
|
BMC Health Serv Res. 2005 Aug 30; 5:57
|
utf-8
|
BMC Health Serv Res
| 2,005 |
10.1186/1472-6963-5-57
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-611615015310.1186/1472-6963-5-61Research ArticleThe use of end-quintile comparisons to identify under-servicing of the poor and over-servicing of the rich: A longitudinal study describing the effect of socioeconomic status on healthcare Brameld Kate J [email protected] C D'Arcy J [email protected] Centre for Health Services Research, School of Population Health, The University of Western Australia, 35 Stirling Highway, Crawley, Western Australia, 60092005 9 9 2005 5 61 61 18 4 2005 9 9 2005 Copyright © 2005 Brameld and Holman; licensee BioMed Central Ltd.2005Brameld and Holman; 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 demonstrate the use of end-quintile comparisons in assessing the effect of socio-economic status on hospital utilisation and outcomes in Western Australia.
Methods
Hospital morbidity records were extracted from the WA Data Linkage System for the period 1994–99, with follow-up to the end of 2000. Multivariate modelling was used to estimate the effect of socio-economic status on hospital admission rates, average and total length of stay (LOS), cumulative incidence of readmission at 30 days and one year, and case fatality at one year.
Results
The study demonstrated higher rate ratios of hospital admission in the more disadvantaged quintiles: rate ratios were 1.31 (95% CI 1.25–1.37) and 1.32 (1.26–1.38) in the first quintile (most disadvantaged) and the second quintile respectively, compared with the fifth quintile (most advantaged). There was a longer total LOS in the most disadvantaged quintile compared with quintile 5 (LOS ratio 1.24; 1.23–1.26). The risk of readmission at 30 days and one year and the risk of death at one year were also greater in those with greater disadvantage: the hazard ratios for quintiles 1:quintile 5 were 1.07 (1.05–1.09), 1.17 (1.16–1.18) and 1.10 (1.07–1.13) respectively. In contradiction to the trends towards higher hospital utilisation and poorer outcomes with increasing social disadvantage, in some MDC's the rate ratio of quintile 1:quintile 2 was less than 1, and quintile 4:quintile 5 was greater than 1. For all surgical admissions the most disadvantaged had a significantly lower admission rate than the second quintile.
Conclusion
This study has shown that the disadvantaged within Western Australia are more intensive users of hospital services but their outcomes following hospitalisation are worse, consistent with their health status. Instances of overuse in the least disadvantaged and under use in the most disadvantaged have also been identified.
==== Body
Background
A recent review of the Australian literature unequivocally showed that the "disadvantaged" had higher mortality rates from most major causes of death, experienced more ill-health and were less likely to act to prevent disease or detect it at an asymptomatic stage [1,2]. Available evidence also suggests that the disadvantaged have experienced more hospital admissions and a higher number of medical consultations [3,4].
This study utilises the unique facility of the Western Australian Linked Data System to provide an overview of the effects of socio-economic status on hospital utilisation and outcomes throughout the State of Western Australia for each major diagnostic category, while adjusting for the effects of age, sex, Aboriginality, comorbidity, geography and possession of health insurance. In addition to providing this general overview of inequalities in hospital utilisation and outcomes in WA, the study also attempts to identify possible examples of under- and over-servicing. Examples of over/under servicing may be indicated, particularly where the procedure is discretionary and where other factors such as disease prevalence, demographic characteristics, medical resources and differing physician practice patterns have been accounted for. [5]
A particular asset of the study is the assignment of socioeconomic status at the level of the Census Collectors' District (CD), which consists of approximately 250 households in urban areas and fewer dwellings in a rural area. This is thus a far more accurate measure than has previously been possible at the postcode level. The data linkage system also enables us to present a greater range of hospital utilisation and outcomes measures and these are given for the complete range of diagnostic categories. In addition we have also been able to adjust for age, sex, aboriginality, accessibility, comorbidity and possession of private health insurance at the individual level.
The study allows areas of inequity to be identified, thus facilitating the development of policies and targeting of resources to address these inequalities.
Method
All linked hospital morbidity and death records were selected from the WA Data Linkage System where a patient had any diagnosis in a specified major diagnostic category (MDC) during the period 1994–1999. For the study of hospital admissions and length of stay (LOS) only records with a diagnosis in the specified MDC between 1994 and 1999 were selected. For the study of cumulative risk of readmission within 30 days and 1 year and case fatality within one year, the patient's first record with a diagnosis of the specified MDC in the period 1994–1999 was selected (the index admission) along with the next admission record with any diagnosis that occurred within a 30 day or 1 year timeframe, as well as any linked death record.
The effect of socio-economic status on hospital admission rates was modelled using Poisson regression. All models were corrected for overdispersion of the data. Multiple linear regression was used to model average LOS of all admissions during the 12 months following index admission and cumulative LOS in the 12 months following index admission. LOS analyses were based on logarithm-transformed data because of skewness in the distribution. Age was squared to improve the model fit. The effects of the study variables on the cumulative risk of readmission within 30 days and 1 year and case fatality within one year were measured using Cox regression. Here the confounding effect of age was modelled using fractional polynomial regression[6]. In all cases, risk adjustment was made for age, sex, Aboriginality, comorbidity, social disadvantage and possession of health insurance. Comorbidity was measured using the Charlson Index, [7,8] which was adapted for use with ICD-10-AM. The allocation of medical and surgical episodes was performed on the basis of DRG. Patient transfers between hospitals were identified and were concatenated into one hospital episode, beginning with the first admission date and ending with the final separation date in the transfer set.
Socio-economic status was measured using the Australian Bureau of Statistics socio-economic indices for areas (SEIFA), specifically the index of relative disadvantage at the level of the census Collector's District (CD). The index of relative socioeconomic disadvantage was one of five SEIFA indices [9]. The indices were created by combining a range of variables using principal component analysis. For the index of relative disadvantage these included qualifications, income, unemployment, type of job, home ownership, one parent families, marital status, car ownership, school leaving age, Aboriginality, number of families per household and fluency in English [9]. In this study SEIFA values were divided into quintiles with 20% of the population in each quintile.
Locational disadvantage was measured using the accessibility/remoteness index of Australia (ARIA),[10]. The ARIA index calculates remoteness based on the distance by road to specified service centres. There are four categories of service centres depending on size, the smallest having a population of 5,000. Every populated locality gets a score between 0 and 12 and these are aggregated into five categories ranging from highly accessible to very remote. For the purpose of this study, the five categories were further collapsed into three; remote, accessible and highly accessible, due to the small number of patients in the less accessible areas.
The measures of locational and social disadvantage were matched to the hospital morbidity records using the CD of residence of the patient as a linkage key. CD's were derived from a geographical point defined by a specific longitude and latitude (a geocode). All patient addresses in the hospital morbidity database were geocoded commencing from records collected in 1993, prior to which the records contained insufficient address information. Even so, approximately 20% of records from 1993 onwards were unable to be geocoded due to incomplete address information. In some sparsely populated areas, CDs for records with missing geocodes could be allocated from the postcodes. In other cases it was possible to allocate a CD from the patient's other linked records, provided that the patient had maintained the same residential postcode and a period of less than five years had elapsed between the admissions. In most of the remaining 16% of records, where a CD was still missing at this stage, the SEIFA and ARIA codes were allocated on the basis of postcode. This left a small proportion of records (1.5%) still without SEIFA and ARIA codes, because not all postcodes were assigned a SEIFA code during the 1996 census. These records were excluded from the analysis.
Possession of private health insurance was allocated on the basis of the payment classification variable having the value of 'private insured'. A patient only required one record with this payment classification to be recorded as privately insured for all analyses in the study. This is based on the assumption that some patients did not always declare or utilise their private health cover, depending on the circumstances of the admission, such as whether emergency or elective treatment was needed; anticipated levels of copayment as a private patient; and whether a privately insured classification was really necessary to secure the patient's choice of doctor. This enabled us to gain an overview of how insurance status affected the patient's entire treatment history, as distinct from the times when they elected to use their private insurance.
Population denominators for the Poisson regression were based on data from the ABS CDATA96 [11]. The proportion of the population with health insurance was taken from the ABS Health Insurance Survey of Australia [12]. The distribution of patients in the population in each comorbidity category was taken as the average distribution across all the MDCs. That is, 72% of patients had no comorbid condition, 23% had one or two comorbidities, 3% had three of four comorbidities and 2% had more than four comorbidities.
The study involved a total of 1,061101 patients admitted over the five year period. Descriptive information about these patients is given in table 1.
Table 1 Descriptive statistics for Patients in the Study
Variable Number Percentage
Agegroup 0–4 188827 17.7
5–9 41760 3.9
10–14 36173 3.4
15–19 57212 5.4
20–24 69732 6.6
25–29 81310 7.7
30–34 82977 7.8
35–39 73647 6.9
40–44 64366 6.1
45–49 62323 5.9
50–54 55034 5.2
55–59 46438 4.4
60–64 42300 4.0
65–69 42244 4.0
70–74 40292 3.8
75+ 76987 7.3
Sex Male 484935 45.7
Female 576166 54.3
No. of comorbidities 0 763993 72.0
1–2 244053 23.0
3–4 31833 3.0
5+ 21222 2.0
SEIFA category 1 – Extreme disadvantage 211371 19.9
2 – disadvantage 243432 22.9
3 – average 225003 21.2
4 – advantaged 184621 17.4
5 – extreme advantage 196674 18.5
ARIA category 1 – Remote 80318 7.6
2 – Accessible 139986 13.2
3 – Highly accessible 840797 79.2
Aborginality Yes 40748 3.8
No 1020353 96.2
Possession of health insurance Yes 348396 32.8
No 712705 67.2
Total persons 1061101
Results
The overall admission rates showed a clear difference between the most and least disadvantaged groups, but there was little difference between the two most disadvantaged quintiles and between the two least advantaged quintiles (table 2). A similar picture was seen for medical admissions, whereas for surgical admissions the most disadvantaged (first quintile) had a significantly lower admission rate than the second quintile.
Table 2 The effect of socio-economic status on hospital utilisation and outcomes
Most disadvantaged Least disadvantaged
Ratio q1:q5 95% CI Ratio q2:q5 95% CI Ratio q3:q5 95% CI Ratio q4:q5 95% CI Baseline – q5
Admission rate – all 1.31 (1.25–1.37) 1.32 (1.26–1.38) 1.20 (1.15–1.25) 0.98 (0.94–1.03) 1.00
Admission rate – medical 1.39 (1.33–1.45) 1.35 (1.29–1.41) 1.20 (1.15–1.26) 0.97 (0.92–1.01) 1.00
Admission rate – surgical 1.12 (1.08–1.15) 1.21 (1.18–1.26) 1.13 (1.09–1.16) 0.98 (0.94–1.01) 1.00
Average LOS 1.04 (1.04–1.05) 1.03 1.03–1.04 1.03 (1.03–1.04) 1.03 (1.02–1.03) 1.00
Total LOS 1.24 (1.23–1.26) 1.14 (1.13–1.16) 1.12 (.10–1.13) 1.08 (1.07–1.10) 1.00
Readmission at 30 days 1.07 (1.05–1.09) 1.05 (1.03–1.06) 1.03 (1.01–1.05) 1.01 (0.99–1.03) 1.00
Readmission at 1 year 1.17 (1.16–1.18) 1.11 (1.10–1.12) 1.07 (1.06–1.08) 1.05 (1.04–1.05) 1.00
Case fatality at 1 year 1.10 (1.07–1.13) 1.05 (1.03–1.08) 1.03 (1.01–1.06) 1.04 (1.01–1.07) 1.00
Socioeconomic status had little effect on average length of stay, but total length of stay decreased with increasing advantage. The risk of readmission within 30 days and 1 year and case fatality within 1 year decreased with increasing social advantage (table 2).
Examination of medical admission rate ratios at the MDC level (table 3) showed that in most cases (9 out of 17 MDCs), those in the most disadvantaged quintile had the highest risk of admission or their risk was similar to those in the second quintile (4 out of 17 MDCs). There were a further four MDCs where the risk in the most disadvantaged was lower than for quintile 2. With regards to the most advantaged quintile, their risk of admission was higher than that for quintile 4 for four out of 17 MDCs, similar to quintile 4 in eight out of 17 MDCs and lower than quintile 4 for five out of 18 MDCs.
Table 3 The effect of socio-economic status on hospital admission rates for medical DRGs
MDC Quintile 1:quintile 2 Quintile 5 :quintile 4
Rate ratio 95% CI Rate ratio 95% CI
1 Nervous 1.05 (1.00 – 1.11) 0.96 (0.90–1.02)
2 Eye* 1.21 (1.15 – 1.28) 1.06 (0.99–1.15)
3 ENTM 0.87 (0.83 – 0.91) 1.17 (1.12–1.24)
4 Respiratory 1.15 (1.10 – 1.20) 0.92 (0.86–0.99)
5 Circulatory 1.02 (0.99 – 1.06) 0.96 (0.92–1.00)
6 Digestive 0.93 (0.89 – 0.96) 1.11 (1.06–1.15)
7 Pancreas*† 1.23 (1.16 – 1.29) 0.90 (0.83–0.96)
8 Musculoskeletal 1.01 (0.97 – 1.05) 1.01 (0.97–1.06)
9 Skin 1.19 (1.14 – 1.26) 1.06 (0.99–1.14)
10 Endocrine*† 1.23 (1.14 – 1.33) 0.93 (0.83–1.05)
11 Kidney*† 1.47 (1.32 – 1.64) 0.65 (0.57–0.75)
12 Male repro† 0.92 (0.85 – 1.00) 1.12 (1.04–1.21)
13 Female repro† 1.18 (1.09 – 1.29) 1.06 (0.96–1.16)
16 Blood 0.93 (0.89 – 0.98) 0.90 (0.85–0.95)
17 Neoplastic 0.89 (0.85 – 0.92) 1.11 (1.07–1.16)
21 Injuries 1.13 (1.08 – 1.18) 1.02 (0.96–1.08)
22 Burns* 1.18 (1.09 – 1.28) 0.90 (0.83–0.98)
Total 1.03 (0.99 – 1.07) 1.03 (0.99–1.09)
Bold text denotes over or underservicing
* questionable convergence of poisson model
† restricted subgroups to allow model convergence
MDC 7 age ≥ 20 years
MDC 10 age ≥ 25 years and 0–2 comorbid conditions
MDC11 age ≥ 30
MDC12 age range 30–64 years and sex male
MDC13 age range 30–64 years and sex female
Analysis of surgical admission rate ratios (table 4) highlighted seven out of 17 MDCs where quintile 1 had a lower risk of admission than quintile 2, six MDCs where the risk was similar and four MDCs where the highest risk was seen in quintile 1. The most advantaged had a higher risk of admission in four out of 17 MDCs, a similar risk in 11 out of 17 MDCs and s lower risk in 2 out of 17 MDCs.
Table 4 The effect of socio-economic status on hospital admission rates for surgical DRGs
MDC Quintile 1:quintile 2 Quintile 5: quintile 4
Rate ratio 95% CI Rate ratio 95% CI
1 Nervous 0.99 (0.95 – 1.03) 0.92 (0.88–0.96)
2 Eye 0.90 (0.88 – 0.93) 1.09 (1.06–1.13)
3 ENTM 0.91 (0.88 – 0.94) 1.02 (0.98–1.05)
4 Respiratory* 1.09 (1.04 – 1.14) 1.03 (0.97–1.09)
5 Circulatory 0.95 (0.91 – 0.98) 0.96 (0.93–1.00)
6 Digestive 0.90 (0.87 – 0.94) 1.01 (0.98–1.05)
7 Pancreas† 1.01 (0.97 – 1.05) 0.97 (0.92–1.02)
8 Musculoskeletal 0.89 (0.86 – 0.92) 1.04 (1.01–1.08)
9 Skin 0.88 (0.84 – 0.92) 1.08 (1.03–1.13)
10 Endocrine† 0.94 (0.88 – 1.01) 1.07 (1.00–1.14)
11 Kidney*† 1.14 (1.06 – 1.23) 0.86 (0.78–0.94)
12 Male repro† 0.95 (0.89 – 1.02) 1.06 (0.98–1.13)
13 Female repro† 0.97 (0.92 – 1.03) 1.00 (0.94–1.05)
16 Blood* 0.97 (0.91 – 1.03) 0.95 (0.89–1.02)
17 Neoplastic* 0.90 (0.85 – 0.96) 1.17 (1.10–1.25)
21 Injuries* 1.07 (1.02 – 1.11) 0.97 (0.92–1.03)
22 Burns* 1.16 (1.07 – 1.26) 0.85 (0.78–0.92)
Total 0.92 (0.89 – 0.95) 1.02 (0.99–1.06)
Bold text denotes over or underservicing
* questionable convergence of poisson model
† restricted subgroups to allow model convergence
End-quintile comparisons of medical and surgical admission rates was used to highlight possible instances of under- and overservicing, whereby the admission rate in the first quintile was at least 10% less than that in the second quintile (ie, rate ratio ≤ 0.90) or the admission rate in the fifth quintile was more than 10% greater than in the fourth quintile (ie. rate ratio ≥ 1.10). Examples of where this occurred are highlighted in tables 3 and 4 for medical and surgical admissions respectively. The most common procedures (those responsible for ≥ 5% of admissions) within the MDCs where the possibility of under- or overservicing was identified are shown in tables 5 and 6.
Table 5 Most common procedures in the MDCs for which underservicing of the socially disadvantaged occurred
MDC ICD-9-CM code Description
2 13.41 Phacoemulsification and aspiration of cataract
13.59 Other extracapsular extraction of lens
3 23.19 Other surgical extraction of teeth
23.13 Surgical extraction of two or more teeth
23.09 Forceps extraction of other tooth
6 47.0 Appendectomy
49.46 Excision of haemorrhoids
53.00 Unilateral repair of inguinal hernia
8 80.6 Excision of semilunar cartilage of knee
81.47 Other repair of knee
9 86.3 Other local excision or destruction of lesion or tissue of skin and subcutaneous tissue
85.21 Local excision of lesion of breast
17 99.25 Chemotherapy
41.31 Biopsy of bone marrow
40.11 Biopsy of lymphatic structures
99.04 Transfusion of packed cells
Table 6 Most common procedures in the MDCs for which overservicing of the socially advantaged occurred
MDC ICD-9-CM code Description
3 23.19 Surgical extraction of teeth
23.13 Surgical extraction of two or more teeth
23.09 Forceps extraction of tooth
6 45.16 Oesophagogastroduodenoscopy with closed biopsy
45.23 colonoscopy
45.42 Colonoscopic polypectomy
45.13 Other endoscopy of small intestine
45.25 Closed biopsy of large intestine
12 63.73 Vasectomy
57.32 Other cystoscopy
60.11 Closed biopsy of prostate
17 99.25 Chemotherapy
41.31 Biopsy of bone marrow
40.11 Biopsy of lymphatic structures
Discussion
Medicare, Australia's universal health insurance scheme was designed to ensure that all Australians had equal access to health care [13]. Equity of access to healthcare in Australia is generally accepted to mean "equal access to equal care for equal need", while recognising that the underprivileged may require more access to more care for the same health problem [14,15]. Numerous studies have shown that those of lower socio-economic status have higher mortality and morbidity rates and their behaviour is more likely to be detrimental to their health; for example, they are more likely to have a poorer diet, be less physically active, drink alcohol to excess and smoke more cigarettes [2]. As a result we would expect them to have a greater requirement for health services.
Our study demonstrated higher hospital admission rates in the disadvantaged. A number of studies have shown that those of low socio-economic status, as measured by various indicators, most commonly income, have a greater risk of hospitalisation than those of high socio-economic status [16-18]. In addition it has been shown that the socio-economic gradient is much greater for medical than surgical admissions as we have also shown (table 1), [16]. However, our study found some anomalies to the general pattern whereby the admission rate was higher in quintile 2 than in quintile 1 (most disadvantaged) and/or the admission rate in the quintile 5 was higher than in quintile 4. This was particularly noticeable for surgical admissions, where the admission rate in quintile 2 was significantly higher than that in quintile 1 (Table 1). This phenomenon is indicative of underservicing in the most disadvantaged and overservicing in the least disadvantaged groups in relation to the segment of the population of nearest socioeconomic status. Further research is required to identify the extent to which this may represent differing rates of access to care or different methods of patient management. It is unlikely to be a result of underlying disease prevalence, which in many cases is higher in the socially disadvantaged. [2]
In some cases, for example chemotherapy, apparent over- or underservicing may have occurred as a result of some patients being treated as day patients (and thus were not included in our hospital morbidity data). Nevertheless, this would still represent a differential pattern of treatment according to socioeconomic group. The cost of surgical extraction of teeth is not covered by Medicare and this is clearly resulting in differential treatment according to socioeconomic status. In other cases, it is likely that those in higher socio-economic groups are better able to negotiate their way through the health system to achieve their desired outcome [19]. While this study has adjusted for possession of private health insurance, the prevalence of which increases with increasing social advantage, more advantaged groups are also better able to pay for procedures in the private sector, as and when necessary [20].
The study showed that hospital admission rates in the most disadvantaged were up to 35% higher than in the most advantaged. In contrast, age-standardised mortality rates for those aged under 65 in Australia during 1998–2000 were 50–90% higher for the most disadvantaged, dependant on age-group and sex [21]. Premature mortality rates have been suggested as the best single indicator of health status that reflects the need for health care and this suggests that the increased admission rate in the most disadvantaged does not match their increased need [22].
The study also demonstrated that the most disadvantaged had a significantly longer total LOS. Canadian data examining all hospitalisations in residents of Winnipeg from 1989–1996, have shown a strong association between length of stay for admissions of 59 days or less and socio-economic status, poorer patients staying in hospital for up to 2.4 times longer than the most affluent patients [18]. A similar result was reported by Epstein et al in a study of nearly 17,000 patients admitted to Massachusetts hospitals during 1987 [23]. They found that patients of lowest socio-economic status had stays of 3–30% longer than those of higher status in 14 out of 15 comparisons when adjusting for age, severity of illness and DRG.
Socio-economic status was also shown to have a significant effect on the outcome measures, risk being greatest in those with greatest disadvantage. However, the effect on the risk of readmission was greater at 1 year than 30 days. This is important because early readmission rates are indicative of possible deficiencies in the process of inpatient care, whereas readmission rates in general may be more an effect of disease progression or the onset of new disease [24]. There were few studies examining the effect of socio-economic status on readmission rates. Weissman et al conducted a study of nearly 12,000 patients adjusted for age, gender, hospital, severity of illness and DRG and found that those of low SES had a greater risk of readmission within 60 days [25].
The effect of socio-economic status on case fatality has mainly been studied following myocardial infarction. Macintyre et al found that the more deprived were more likely to die within 30 days of their myocardial infarction. Alter et al found a similar relationship with one year mortality the Salomaa et al also found that case fatality at 28 days and one year was highest in those with low income and education, consistent with our results [26-28].
Increased use of hospital services by the more disadvantaged in our population reflect their greater health need and, in many cases, also a greater severity of illness [23,29]. Weissman et al suggested that greater difficulty in accessing ambulatory care post-discharge, inability to afford recommended therapies and possible non-compliance or misunderstanding of physicians orders may increase the risk of readmissions [25].
Conclusion
This study has shown that the socially disadvantaged within Western Australia are more intensive users of health services and their outcomes following hospitalisation are worse. Comparison of hospital admission rates and premature mortality rates suggest that the increased admission rates in the most disadvantaged are not sufficient to account for their increased need. Some examples of overuse in the least disadvantaged and under use in the most disadvantaged were also identified. These factors together with the discrepancy in the gradient between medical and surgical admissions are suggestive of inequity in treatment between socio-economic groups, specifically the greatest and the least disadvantaged.
The WA data linkage system provides an effective mechanism for ongoing monitoring of equity in hospital utilisation and outcomes. Further research is required to identify the need for health services according to socio-economic groups and to identify how services can be made more easily accessible for these groups.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KJB participated in the conception and design of the study, performed the analysis and drafted the paper. CDJH participated in the conception and design of the study and in revising the paper.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors wish to acknowledge all staff who maintain the Western Australian Hospital Morbidity Data System and the Data Linkage System on which this study was based. The study was funded by the National Health and Medical Research Council, Australia. Grant No. 139071.
==== Refs
Turrell G Oldenburg B McGuffog I Dent R Socioeconomic determinants of health: towards a national research program and a policy and intervention agenda 1999 Canberra: Queensland University of technology
Turrell G Mathers C Socioeconomic status and health in Australia Med J Aust 2000 172 434 438 10870537
Taylor R Quine S Lyle D Bilton A Socioeconomic correlates of mortality and morbidity in Sydney, 1985–88 Aust J Public Health 1992 16 305 314 1482725
Wiggers J Sanson-Fisher R Halpin S Prevalence and frequency of health service use: Associations with occupational prestige and educational attainment Aust J Public Health 1995 19 512 519 8713203
Gittelsohn A Powe N Small area variation in health care delivery in Maryland Health Serv Res 1995 30 295 317 7782218
Royston P Altman D Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling (with discussion) Appl Stat 1994 43 429 467
Charlson M Pompei P Ales K Mackenzie C A new method of classifying prognostic comorbidity in longitudinal studies: development and validation J Chronic Dis 1987 40 373 383 3558716 10.1016/0021-9681(87)90171-8
Romano PS Roos LL Jollis JG Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives J Clin Epidemiol 1993 46 1075 9 discussion 1081–90 8410092 10.1016/0895-4356(93)90103-8
ABS 1996 Census of Population and Housing. Socio-economic indexes for areas 1998 Canberra: Australian Bureau of Statistics
CDHAC Measuring Remoteness: Accessibility/remoteness index of Australia (ARIA) 2001 Canberra: Commonwealth Dept of Health and Aged Care
ABS 1996 Census of Population and Housing. CDATA96 1997 Canberra: Australian Bureau of Statistics
ABS Health Insurance Survey, Australia 1999 Canberra: Australian Bureau of Statistics
Duckett S The Australian Health Care System 2004 2 Melbourne: Oxford University Press
Mooney G Economics, medicine and health care 2003 3 London: Prentice Hall
Braverman P Gruskin S Poverty, equity, human rights, and health Bulletin World Health Organ 2003 81 539 545
Keskimaki I How did Finland's economic recession in the early 1990s affect socio-economic equity in the use of hospital care? Soc Sci Med 2003 56 1517 1530 12614702 10.1016/S0277-9536(02)00153-3
Taylor R Chey T Bauman A Webster I Socio-economic, migrant and geographic differentials in coronary heart disease occurence in New South Wales Aust N Z J Public Health 1999 23 20 26 10083685
Carriere K Roos L Dover D Across time and space: variations in hospital use during Canadian health reform Health Serv Res 2000 35 467 87 10857472
Roos N Mustard C Variation in health and health care use by socioeconomic status in Winnipeg, Canada: Does the system work well? Yes and no Milbank Q 1997 75 89 111 9063301 10.1111/1468-0009.00045
ABS Private Health Insurance 2003 Canberra: Australian Bureau of Statistics
Draper G Turrell G Oldenburg B Health inequalities in Australia: Mortality 2004 Canberra: Queensland Institute of Technology and the Australian Institute of Health and Welfare
Carstairs V Morris R Deprivation and Health in Scotland 1991 Aberdeen: Aberdeen University Press
Epstein A Stern R Weissman J Do the poor cost more? A multihospital study of patients' socioeconomic status and use of hospital resources N Engl J Med 1990 322 1122 8 2108331
Ashton CM Wray NP A conceptual framework for the study of early readmission as an indicator of quality of care Soc Sci Med 1996 43 1533 41 8961397 10.1016/S0277-9536(96)00049-4
Weissman J Stern R Epstein A The Impact of Patient Socioeconomic Status and Other Social Factors on Readmission: a Prospective Study in Four Massachusetts Hospitals Inquiry 1994 31 163 172 8021022
Macintyre K Stewart S Chalmers J Pell J Finlayson A Boyd J Relation between socioeconomic deprivation and death from a first myocardial infarction in Scotland: population based analysis BMJ 2001 322 1152 53 11348909 10.1136/bmj.322.7295.1152
Salomaa V Niemela M Miettinen H Ketonen M Immonen-Raija P Koskinene S Relationship of socioeconomic status to the incidence and prehospital, 28-Day, and 1-Year mortality rates of acute coronary events in the FINMONICA myocardial infarction register study Circulation 2000 101 1913 1918 10779456
Alter D CD N Austin P Tu J Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction N Engl J Med 1999 341 1359 67 10536129 10.1056/NEJM199910283411806
Eachus J Chan P Pearson N Propper C Davey Smith G An additional dimension to health inequalities: disease severity and socioeconomic position J Epidemiol Community Health 1999 53 603 611 10616672
|
16150153
|
PMC1236924
|
CC BY
|
2021-01-04 16:31:50
|
no
|
BMC Health Serv Res. 2005 Sep 9; 5:61
|
utf-8
|
BMC Health Serv Res
| 2,005 |
10.1186/1472-6963-5-61
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-611615015310.1186/1472-6963-5-61Research ArticleThe use of end-quintile comparisons to identify under-servicing of the poor and over-servicing of the rich: A longitudinal study describing the effect of socioeconomic status on healthcare Brameld Kate J [email protected] C D'Arcy J [email protected] Centre for Health Services Research, School of Population Health, The University of Western Australia, 35 Stirling Highway, Crawley, Western Australia, 60092005 9 9 2005 5 61 61 18 4 2005 9 9 2005 Copyright © 2005 Brameld and Holman; licensee BioMed Central Ltd.2005Brameld and Holman; 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 demonstrate the use of end-quintile comparisons in assessing the effect of socio-economic status on hospital utilisation and outcomes in Western Australia.
Methods
Hospital morbidity records were extracted from the WA Data Linkage System for the period 1994–99, with follow-up to the end of 2000. Multivariate modelling was used to estimate the effect of socio-economic status on hospital admission rates, average and total length of stay (LOS), cumulative incidence of readmission at 30 days and one year, and case fatality at one year.
Results
The study demonstrated higher rate ratios of hospital admission in the more disadvantaged quintiles: rate ratios were 1.31 (95% CI 1.25–1.37) and 1.32 (1.26–1.38) in the first quintile (most disadvantaged) and the second quintile respectively, compared with the fifth quintile (most advantaged). There was a longer total LOS in the most disadvantaged quintile compared with quintile 5 (LOS ratio 1.24; 1.23–1.26). The risk of readmission at 30 days and one year and the risk of death at one year were also greater in those with greater disadvantage: the hazard ratios for quintiles 1:quintile 5 were 1.07 (1.05–1.09), 1.17 (1.16–1.18) and 1.10 (1.07–1.13) respectively. In contradiction to the trends towards higher hospital utilisation and poorer outcomes with increasing social disadvantage, in some MDC's the rate ratio of quintile 1:quintile 2 was less than 1, and quintile 4:quintile 5 was greater than 1. For all surgical admissions the most disadvantaged had a significantly lower admission rate than the second quintile.
Conclusion
This study has shown that the disadvantaged within Western Australia are more intensive users of hospital services but their outcomes following hospitalisation are worse, consistent with their health status. Instances of overuse in the least disadvantaged and under use in the most disadvantaged have also been identified.
==== Body
Background
A recent review of the Australian literature unequivocally showed that the "disadvantaged" had higher mortality rates from most major causes of death, experienced more ill-health and were less likely to act to prevent disease or detect it at an asymptomatic stage [1,2]. Available evidence also suggests that the disadvantaged have experienced more hospital admissions and a higher number of medical consultations [3,4].
This study utilises the unique facility of the Western Australian Linked Data System to provide an overview of the effects of socio-economic status on hospital utilisation and outcomes throughout the State of Western Australia for each major diagnostic category, while adjusting for the effects of age, sex, Aboriginality, comorbidity, geography and possession of health insurance. In addition to providing this general overview of inequalities in hospital utilisation and outcomes in WA, the study also attempts to identify possible examples of under- and over-servicing. Examples of over/under servicing may be indicated, particularly where the procedure is discretionary and where other factors such as disease prevalence, demographic characteristics, medical resources and differing physician practice patterns have been accounted for. [5]
A particular asset of the study is the assignment of socioeconomic status at the level of the Census Collectors' District (CD), which consists of approximately 250 households in urban areas and fewer dwellings in a rural area. This is thus a far more accurate measure than has previously been possible at the postcode level. The data linkage system also enables us to present a greater range of hospital utilisation and outcomes measures and these are given for the complete range of diagnostic categories. In addition we have also been able to adjust for age, sex, aboriginality, accessibility, comorbidity and possession of private health insurance at the individual level.
The study allows areas of inequity to be identified, thus facilitating the development of policies and targeting of resources to address these inequalities.
Method
All linked hospital morbidity and death records were selected from the WA Data Linkage System where a patient had any diagnosis in a specified major diagnostic category (MDC) during the period 1994–1999. For the study of hospital admissions and length of stay (LOS) only records with a diagnosis in the specified MDC between 1994 and 1999 were selected. For the study of cumulative risk of readmission within 30 days and 1 year and case fatality within one year, the patient's first record with a diagnosis of the specified MDC in the period 1994–1999 was selected (the index admission) along with the next admission record with any diagnosis that occurred within a 30 day or 1 year timeframe, as well as any linked death record.
The effect of socio-economic status on hospital admission rates was modelled using Poisson regression. All models were corrected for overdispersion of the data. Multiple linear regression was used to model average LOS of all admissions during the 12 months following index admission and cumulative LOS in the 12 months following index admission. LOS analyses were based on logarithm-transformed data because of skewness in the distribution. Age was squared to improve the model fit. The effects of the study variables on the cumulative risk of readmission within 30 days and 1 year and case fatality within one year were measured using Cox regression. Here the confounding effect of age was modelled using fractional polynomial regression[6]. In all cases, risk adjustment was made for age, sex, Aboriginality, comorbidity, social disadvantage and possession of health insurance. Comorbidity was measured using the Charlson Index, [7,8] which was adapted for use with ICD-10-AM. The allocation of medical and surgical episodes was performed on the basis of DRG. Patient transfers between hospitals were identified and were concatenated into one hospital episode, beginning with the first admission date and ending with the final separation date in the transfer set.
Socio-economic status was measured using the Australian Bureau of Statistics socio-economic indices for areas (SEIFA), specifically the index of relative disadvantage at the level of the census Collector's District (CD). The index of relative socioeconomic disadvantage was one of five SEIFA indices [9]. The indices were created by combining a range of variables using principal component analysis. For the index of relative disadvantage these included qualifications, income, unemployment, type of job, home ownership, one parent families, marital status, car ownership, school leaving age, Aboriginality, number of families per household and fluency in English [9]. In this study SEIFA values were divided into quintiles with 20% of the population in each quintile.
Locational disadvantage was measured using the accessibility/remoteness index of Australia (ARIA),[10]. The ARIA index calculates remoteness based on the distance by road to specified service centres. There are four categories of service centres depending on size, the smallest having a population of 5,000. Every populated locality gets a score between 0 and 12 and these are aggregated into five categories ranging from highly accessible to very remote. For the purpose of this study, the five categories were further collapsed into three; remote, accessible and highly accessible, due to the small number of patients in the less accessible areas.
The measures of locational and social disadvantage were matched to the hospital morbidity records using the CD of residence of the patient as a linkage key. CD's were derived from a geographical point defined by a specific longitude and latitude (a geocode). All patient addresses in the hospital morbidity database were geocoded commencing from records collected in 1993, prior to which the records contained insufficient address information. Even so, approximately 20% of records from 1993 onwards were unable to be geocoded due to incomplete address information. In some sparsely populated areas, CDs for records with missing geocodes could be allocated from the postcodes. In other cases it was possible to allocate a CD from the patient's other linked records, provided that the patient had maintained the same residential postcode and a period of less than five years had elapsed between the admissions. In most of the remaining 16% of records, where a CD was still missing at this stage, the SEIFA and ARIA codes were allocated on the basis of postcode. This left a small proportion of records (1.5%) still without SEIFA and ARIA codes, because not all postcodes were assigned a SEIFA code during the 1996 census. These records were excluded from the analysis.
Possession of private health insurance was allocated on the basis of the payment classification variable having the value of 'private insured'. A patient only required one record with this payment classification to be recorded as privately insured for all analyses in the study. This is based on the assumption that some patients did not always declare or utilise their private health cover, depending on the circumstances of the admission, such as whether emergency or elective treatment was needed; anticipated levels of copayment as a private patient; and whether a privately insured classification was really necessary to secure the patient's choice of doctor. This enabled us to gain an overview of how insurance status affected the patient's entire treatment history, as distinct from the times when they elected to use their private insurance.
Population denominators for the Poisson regression were based on data from the ABS CDATA96 [11]. The proportion of the population with health insurance was taken from the ABS Health Insurance Survey of Australia [12]. The distribution of patients in the population in each comorbidity category was taken as the average distribution across all the MDCs. That is, 72% of patients had no comorbid condition, 23% had one or two comorbidities, 3% had three of four comorbidities and 2% had more than four comorbidities.
The study involved a total of 1,061101 patients admitted over the five year period. Descriptive information about these patients is given in table 1.
Table 1 Descriptive statistics for Patients in the Study
Variable Number Percentage
Agegroup 0–4 188827 17.7
5–9 41760 3.9
10–14 36173 3.4
15–19 57212 5.4
20–24 69732 6.6
25–29 81310 7.7
30–34 82977 7.8
35–39 73647 6.9
40–44 64366 6.1
45–49 62323 5.9
50–54 55034 5.2
55–59 46438 4.4
60–64 42300 4.0
65–69 42244 4.0
70–74 40292 3.8
75+ 76987 7.3
Sex Male 484935 45.7
Female 576166 54.3
No. of comorbidities 0 763993 72.0
1–2 244053 23.0
3–4 31833 3.0
5+ 21222 2.0
SEIFA category 1 – Extreme disadvantage 211371 19.9
2 – disadvantage 243432 22.9
3 – average 225003 21.2
4 – advantaged 184621 17.4
5 – extreme advantage 196674 18.5
ARIA category 1 – Remote 80318 7.6
2 – Accessible 139986 13.2
3 – Highly accessible 840797 79.2
Aborginality Yes 40748 3.8
No 1020353 96.2
Possession of health insurance Yes 348396 32.8
No 712705 67.2
Total persons 1061101
Results
The overall admission rates showed a clear difference between the most and least disadvantaged groups, but there was little difference between the two most disadvantaged quintiles and between the two least advantaged quintiles (table 2). A similar picture was seen for medical admissions, whereas for surgical admissions the most disadvantaged (first quintile) had a significantly lower admission rate than the second quintile.
Table 2 The effect of socio-economic status on hospital utilisation and outcomes
Most disadvantaged Least disadvantaged
Ratio q1:q5 95% CI Ratio q2:q5 95% CI Ratio q3:q5 95% CI Ratio q4:q5 95% CI Baseline – q5
Admission rate – all 1.31 (1.25–1.37) 1.32 (1.26–1.38) 1.20 (1.15–1.25) 0.98 (0.94–1.03) 1.00
Admission rate – medical 1.39 (1.33–1.45) 1.35 (1.29–1.41) 1.20 (1.15–1.26) 0.97 (0.92–1.01) 1.00
Admission rate – surgical 1.12 (1.08–1.15) 1.21 (1.18–1.26) 1.13 (1.09–1.16) 0.98 (0.94–1.01) 1.00
Average LOS 1.04 (1.04–1.05) 1.03 1.03–1.04 1.03 (1.03–1.04) 1.03 (1.02–1.03) 1.00
Total LOS 1.24 (1.23–1.26) 1.14 (1.13–1.16) 1.12 (.10–1.13) 1.08 (1.07–1.10) 1.00
Readmission at 30 days 1.07 (1.05–1.09) 1.05 (1.03–1.06) 1.03 (1.01–1.05) 1.01 (0.99–1.03) 1.00
Readmission at 1 year 1.17 (1.16–1.18) 1.11 (1.10–1.12) 1.07 (1.06–1.08) 1.05 (1.04–1.05) 1.00
Case fatality at 1 year 1.10 (1.07–1.13) 1.05 (1.03–1.08) 1.03 (1.01–1.06) 1.04 (1.01–1.07) 1.00
Socioeconomic status had little effect on average length of stay, but total length of stay decreased with increasing advantage. The risk of readmission within 30 days and 1 year and case fatality within 1 year decreased with increasing social advantage (table 2).
Examination of medical admission rate ratios at the MDC level (table 3) showed that in most cases (9 out of 17 MDCs), those in the most disadvantaged quintile had the highest risk of admission or their risk was similar to those in the second quintile (4 out of 17 MDCs). There were a further four MDCs where the risk in the most disadvantaged was lower than for quintile 2. With regards to the most advantaged quintile, their risk of admission was higher than that for quintile 4 for four out of 17 MDCs, similar to quintile 4 in eight out of 17 MDCs and lower than quintile 4 for five out of 18 MDCs.
Table 3 The effect of socio-economic status on hospital admission rates for medical DRGs
MDC Quintile 1:quintile 2 Quintile 5 :quintile 4
Rate ratio 95% CI Rate ratio 95% CI
1 Nervous 1.05 (1.00 – 1.11) 0.96 (0.90–1.02)
2 Eye* 1.21 (1.15 – 1.28) 1.06 (0.99–1.15)
3 ENTM 0.87 (0.83 – 0.91) 1.17 (1.12–1.24)
4 Respiratory 1.15 (1.10 – 1.20) 0.92 (0.86–0.99)
5 Circulatory 1.02 (0.99 – 1.06) 0.96 (0.92–1.00)
6 Digestive 0.93 (0.89 – 0.96) 1.11 (1.06–1.15)
7 Pancreas*† 1.23 (1.16 – 1.29) 0.90 (0.83–0.96)
8 Musculoskeletal 1.01 (0.97 – 1.05) 1.01 (0.97–1.06)
9 Skin 1.19 (1.14 – 1.26) 1.06 (0.99–1.14)
10 Endocrine*† 1.23 (1.14 – 1.33) 0.93 (0.83–1.05)
11 Kidney*† 1.47 (1.32 – 1.64) 0.65 (0.57–0.75)
12 Male repro† 0.92 (0.85 – 1.00) 1.12 (1.04–1.21)
13 Female repro† 1.18 (1.09 – 1.29) 1.06 (0.96–1.16)
16 Blood 0.93 (0.89 – 0.98) 0.90 (0.85–0.95)
17 Neoplastic 0.89 (0.85 – 0.92) 1.11 (1.07–1.16)
21 Injuries 1.13 (1.08 – 1.18) 1.02 (0.96–1.08)
22 Burns* 1.18 (1.09 – 1.28) 0.90 (0.83–0.98)
Total 1.03 (0.99 – 1.07) 1.03 (0.99–1.09)
Bold text denotes over or underservicing
* questionable convergence of poisson model
† restricted subgroups to allow model convergence
MDC 7 age ≥ 20 years
MDC 10 age ≥ 25 years and 0–2 comorbid conditions
MDC11 age ≥ 30
MDC12 age range 30–64 years and sex male
MDC13 age range 30–64 years and sex female
Analysis of surgical admission rate ratios (table 4) highlighted seven out of 17 MDCs where quintile 1 had a lower risk of admission than quintile 2, six MDCs where the risk was similar and four MDCs where the highest risk was seen in quintile 1. The most advantaged had a higher risk of admission in four out of 17 MDCs, a similar risk in 11 out of 17 MDCs and s lower risk in 2 out of 17 MDCs.
Table 4 The effect of socio-economic status on hospital admission rates for surgical DRGs
MDC Quintile 1:quintile 2 Quintile 5: quintile 4
Rate ratio 95% CI Rate ratio 95% CI
1 Nervous 0.99 (0.95 – 1.03) 0.92 (0.88–0.96)
2 Eye 0.90 (0.88 – 0.93) 1.09 (1.06–1.13)
3 ENTM 0.91 (0.88 – 0.94) 1.02 (0.98–1.05)
4 Respiratory* 1.09 (1.04 – 1.14) 1.03 (0.97–1.09)
5 Circulatory 0.95 (0.91 – 0.98) 0.96 (0.93–1.00)
6 Digestive 0.90 (0.87 – 0.94) 1.01 (0.98–1.05)
7 Pancreas† 1.01 (0.97 – 1.05) 0.97 (0.92–1.02)
8 Musculoskeletal 0.89 (0.86 – 0.92) 1.04 (1.01–1.08)
9 Skin 0.88 (0.84 – 0.92) 1.08 (1.03–1.13)
10 Endocrine† 0.94 (0.88 – 1.01) 1.07 (1.00–1.14)
11 Kidney*† 1.14 (1.06 – 1.23) 0.86 (0.78–0.94)
12 Male repro† 0.95 (0.89 – 1.02) 1.06 (0.98–1.13)
13 Female repro† 0.97 (0.92 – 1.03) 1.00 (0.94–1.05)
16 Blood* 0.97 (0.91 – 1.03) 0.95 (0.89–1.02)
17 Neoplastic* 0.90 (0.85 – 0.96) 1.17 (1.10–1.25)
21 Injuries* 1.07 (1.02 – 1.11) 0.97 (0.92–1.03)
22 Burns* 1.16 (1.07 – 1.26) 0.85 (0.78–0.92)
Total 0.92 (0.89 – 0.95) 1.02 (0.99–1.06)
Bold text denotes over or underservicing
* questionable convergence of poisson model
† restricted subgroups to allow model convergence
End-quintile comparisons of medical and surgical admission rates was used to highlight possible instances of under- and overservicing, whereby the admission rate in the first quintile was at least 10% less than that in the second quintile (ie, rate ratio ≤ 0.90) or the admission rate in the fifth quintile was more than 10% greater than in the fourth quintile (ie. rate ratio ≥ 1.10). Examples of where this occurred are highlighted in tables 3 and 4 for medical and surgical admissions respectively. The most common procedures (those responsible for ≥ 5% of admissions) within the MDCs where the possibility of under- or overservicing was identified are shown in tables 5 and 6.
Table 5 Most common procedures in the MDCs for which underservicing of the socially disadvantaged occurred
MDC ICD-9-CM code Description
2 13.41 Phacoemulsification and aspiration of cataract
13.59 Other extracapsular extraction of lens
3 23.19 Other surgical extraction of teeth
23.13 Surgical extraction of two or more teeth
23.09 Forceps extraction of other tooth
6 47.0 Appendectomy
49.46 Excision of haemorrhoids
53.00 Unilateral repair of inguinal hernia
8 80.6 Excision of semilunar cartilage of knee
81.47 Other repair of knee
9 86.3 Other local excision or destruction of lesion or tissue of skin and subcutaneous tissue
85.21 Local excision of lesion of breast
17 99.25 Chemotherapy
41.31 Biopsy of bone marrow
40.11 Biopsy of lymphatic structures
99.04 Transfusion of packed cells
Table 6 Most common procedures in the MDCs for which overservicing of the socially advantaged occurred
MDC ICD-9-CM code Description
3 23.19 Surgical extraction of teeth
23.13 Surgical extraction of two or more teeth
23.09 Forceps extraction of tooth
6 45.16 Oesophagogastroduodenoscopy with closed biopsy
45.23 colonoscopy
45.42 Colonoscopic polypectomy
45.13 Other endoscopy of small intestine
45.25 Closed biopsy of large intestine
12 63.73 Vasectomy
57.32 Other cystoscopy
60.11 Closed biopsy of prostate
17 99.25 Chemotherapy
41.31 Biopsy of bone marrow
40.11 Biopsy of lymphatic structures
Discussion
Medicare, Australia's universal health insurance scheme was designed to ensure that all Australians had equal access to health care [13]. Equity of access to healthcare in Australia is generally accepted to mean "equal access to equal care for equal need", while recognising that the underprivileged may require more access to more care for the same health problem [14,15]. Numerous studies have shown that those of lower socio-economic status have higher mortality and morbidity rates and their behaviour is more likely to be detrimental to their health; for example, they are more likely to have a poorer diet, be less physically active, drink alcohol to excess and smoke more cigarettes [2]. As a result we would expect them to have a greater requirement for health services.
Our study demonstrated higher hospital admission rates in the disadvantaged. A number of studies have shown that those of low socio-economic status, as measured by various indicators, most commonly income, have a greater risk of hospitalisation than those of high socio-economic status [16-18]. In addition it has been shown that the socio-economic gradient is much greater for medical than surgical admissions as we have also shown (table 1), [16]. However, our study found some anomalies to the general pattern whereby the admission rate was higher in quintile 2 than in quintile 1 (most disadvantaged) and/or the admission rate in the quintile 5 was higher than in quintile 4. This was particularly noticeable for surgical admissions, where the admission rate in quintile 2 was significantly higher than that in quintile 1 (Table 1). This phenomenon is indicative of underservicing in the most disadvantaged and overservicing in the least disadvantaged groups in relation to the segment of the population of nearest socioeconomic status. Further research is required to identify the extent to which this may represent differing rates of access to care or different methods of patient management. It is unlikely to be a result of underlying disease prevalence, which in many cases is higher in the socially disadvantaged. [2]
In some cases, for example chemotherapy, apparent over- or underservicing may have occurred as a result of some patients being treated as day patients (and thus were not included in our hospital morbidity data). Nevertheless, this would still represent a differential pattern of treatment according to socioeconomic group. The cost of surgical extraction of teeth is not covered by Medicare and this is clearly resulting in differential treatment according to socioeconomic status. In other cases, it is likely that those in higher socio-economic groups are better able to negotiate their way through the health system to achieve their desired outcome [19]. While this study has adjusted for possession of private health insurance, the prevalence of which increases with increasing social advantage, more advantaged groups are also better able to pay for procedures in the private sector, as and when necessary [20].
The study showed that hospital admission rates in the most disadvantaged were up to 35% higher than in the most advantaged. In contrast, age-standardised mortality rates for those aged under 65 in Australia during 1998–2000 were 50–90% higher for the most disadvantaged, dependant on age-group and sex [21]. Premature mortality rates have been suggested as the best single indicator of health status that reflects the need for health care and this suggests that the increased admission rate in the most disadvantaged does not match their increased need [22].
The study also demonstrated that the most disadvantaged had a significantly longer total LOS. Canadian data examining all hospitalisations in residents of Winnipeg from 1989–1996, have shown a strong association between length of stay for admissions of 59 days or less and socio-economic status, poorer patients staying in hospital for up to 2.4 times longer than the most affluent patients [18]. A similar result was reported by Epstein et al in a study of nearly 17,000 patients admitted to Massachusetts hospitals during 1987 [23]. They found that patients of lowest socio-economic status had stays of 3–30% longer than those of higher status in 14 out of 15 comparisons when adjusting for age, severity of illness and DRG.
Socio-economic status was also shown to have a significant effect on the outcome measures, risk being greatest in those with greatest disadvantage. However, the effect on the risk of readmission was greater at 1 year than 30 days. This is important because early readmission rates are indicative of possible deficiencies in the process of inpatient care, whereas readmission rates in general may be more an effect of disease progression or the onset of new disease [24]. There were few studies examining the effect of socio-economic status on readmission rates. Weissman et al conducted a study of nearly 12,000 patients adjusted for age, gender, hospital, severity of illness and DRG and found that those of low SES had a greater risk of readmission within 60 days [25].
The effect of socio-economic status on case fatality has mainly been studied following myocardial infarction. Macintyre et al found that the more deprived were more likely to die within 30 days of their myocardial infarction. Alter et al found a similar relationship with one year mortality the Salomaa et al also found that case fatality at 28 days and one year was highest in those with low income and education, consistent with our results [26-28].
Increased use of hospital services by the more disadvantaged in our population reflect their greater health need and, in many cases, also a greater severity of illness [23,29]. Weissman et al suggested that greater difficulty in accessing ambulatory care post-discharge, inability to afford recommended therapies and possible non-compliance or misunderstanding of physicians orders may increase the risk of readmissions [25].
Conclusion
This study has shown that the socially disadvantaged within Western Australia are more intensive users of health services and their outcomes following hospitalisation are worse. Comparison of hospital admission rates and premature mortality rates suggest that the increased admission rates in the most disadvantaged are not sufficient to account for their increased need. Some examples of overuse in the least disadvantaged and under use in the most disadvantaged were also identified. These factors together with the discrepancy in the gradient between medical and surgical admissions are suggestive of inequity in treatment between socio-economic groups, specifically the greatest and the least disadvantaged.
The WA data linkage system provides an effective mechanism for ongoing monitoring of equity in hospital utilisation and outcomes. Further research is required to identify the need for health services according to socio-economic groups and to identify how services can be made more easily accessible for these groups.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KJB participated in the conception and design of the study, performed the analysis and drafted the paper. CDJH participated in the conception and design of the study and in revising the paper.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors wish to acknowledge all staff who maintain the Western Australian Hospital Morbidity Data System and the Data Linkage System on which this study was based. The study was funded by the National Health and Medical Research Council, Australia. Grant No. 139071.
==== Refs
Turrell G Oldenburg B McGuffog I Dent R Socioeconomic determinants of health: towards a national research program and a policy and intervention agenda 1999 Canberra: Queensland University of technology
Turrell G Mathers C Socioeconomic status and health in Australia Med J Aust 2000 172 434 438 10870537
Taylor R Quine S Lyle D Bilton A Socioeconomic correlates of mortality and morbidity in Sydney, 1985–88 Aust J Public Health 1992 16 305 314 1482725
Wiggers J Sanson-Fisher R Halpin S Prevalence and frequency of health service use: Associations with occupational prestige and educational attainment Aust J Public Health 1995 19 512 519 8713203
Gittelsohn A Powe N Small area variation in health care delivery in Maryland Health Serv Res 1995 30 295 317 7782218
Royston P Altman D Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling (with discussion) Appl Stat 1994 43 429 467
Charlson M Pompei P Ales K Mackenzie C A new method of classifying prognostic comorbidity in longitudinal studies: development and validation J Chronic Dis 1987 40 373 383 3558716 10.1016/0021-9681(87)90171-8
Romano PS Roos LL Jollis JG Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives J Clin Epidemiol 1993 46 1075 9 discussion 1081–90 8410092 10.1016/0895-4356(93)90103-8
ABS 1996 Census of Population and Housing. Socio-economic indexes for areas 1998 Canberra: Australian Bureau of Statistics
CDHAC Measuring Remoteness: Accessibility/remoteness index of Australia (ARIA) 2001 Canberra: Commonwealth Dept of Health and Aged Care
ABS 1996 Census of Population and Housing. CDATA96 1997 Canberra: Australian Bureau of Statistics
ABS Health Insurance Survey, Australia 1999 Canberra: Australian Bureau of Statistics
Duckett S The Australian Health Care System 2004 2 Melbourne: Oxford University Press
Mooney G Economics, medicine and health care 2003 3 London: Prentice Hall
Braverman P Gruskin S Poverty, equity, human rights, and health Bulletin World Health Organ 2003 81 539 545
Keskimaki I How did Finland's economic recession in the early 1990s affect socio-economic equity in the use of hospital care? Soc Sci Med 2003 56 1517 1530 12614702 10.1016/S0277-9536(02)00153-3
Taylor R Chey T Bauman A Webster I Socio-economic, migrant and geographic differentials in coronary heart disease occurence in New South Wales Aust N Z J Public Health 1999 23 20 26 10083685
Carriere K Roos L Dover D Across time and space: variations in hospital use during Canadian health reform Health Serv Res 2000 35 467 87 10857472
Roos N Mustard C Variation in health and health care use by socioeconomic status in Winnipeg, Canada: Does the system work well? Yes and no Milbank Q 1997 75 89 111 9063301 10.1111/1468-0009.00045
ABS Private Health Insurance 2003 Canberra: Australian Bureau of Statistics
Draper G Turrell G Oldenburg B Health inequalities in Australia: Mortality 2004 Canberra: Queensland Institute of Technology and the Australian Institute of Health and Welfare
Carstairs V Morris R Deprivation and Health in Scotland 1991 Aberdeen: Aberdeen University Press
Epstein A Stern R Weissman J Do the poor cost more? A multihospital study of patients' socioeconomic status and use of hospital resources N Engl J Med 1990 322 1122 8 2108331
Ashton CM Wray NP A conceptual framework for the study of early readmission as an indicator of quality of care Soc Sci Med 1996 43 1533 41 8961397 10.1016/S0277-9536(96)00049-4
Weissman J Stern R Epstein A The Impact of Patient Socioeconomic Status and Other Social Factors on Readmission: a Prospective Study in Four Massachusetts Hospitals Inquiry 1994 31 163 172 8021022
Macintyre K Stewart S Chalmers J Pell J Finlayson A Boyd J Relation between socioeconomic deprivation and death from a first myocardial infarction in Scotland: population based analysis BMJ 2001 322 1152 53 11348909 10.1136/bmj.322.7295.1152
Salomaa V Niemela M Miettinen H Ketonen M Immonen-Raija P Koskinene S Relationship of socioeconomic status to the incidence and prehospital, 28-Day, and 1-Year mortality rates of acute coronary events in the FINMONICA myocardial infarction register study Circulation 2000 101 1913 1918 10779456
Alter D CD N Austin P Tu J Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction N Engl J Med 1999 341 1359 67 10536129 10.1056/NEJM199910283411806
Eachus J Chan P Pearson N Propper C Davey Smith G An additional dimension to health inequalities: disease severity and socioeconomic position J Epidemiol Community Health 1999 53 603 611 10616672
|
16144543
|
PMC1236925
|
CC BY
|
2021-01-04 16:28:15
|
no
|
BMC Infect Dis. 2005 Sep 6; 5:68
|
latin-1
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-68
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-691615689210.1186/1471-2334-5-69Research ArticleDifferent patterns of HIV-1 DNA after therapy discontinuation Re Maria Carla [email protected] Francesca [email protected] Laura [email protected] Pasqua [email protected] Florio [email protected] Davide [email protected] Department of Clinical and Experimental Medicine, Section of Microbiology, University of Bologna, Via Massarenti 9-40138 Bologna, Italy2 Department of Infectious Diseases, St Anna Hospital, Corso Giovecca, 203-44100 Ferrara, Italy2005 12 9 2005 5 69 69 29 4 2005 12 9 2005 Copyright © 2005 Carla Re et al; licensee BioMed Central Ltd.2005Carla Re 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
By persisting in infected cells for a long period of time, proviral HIV-1 DNA can represent an alternative viral marker to RNA viral load during the follow-up of HIV-1 infected individuals. In the present study sequential blood samples of 10 patients under antiretroviral treatment from 1997 with two NRTIs, who refused to continue any antiviral regimen, were analyzed for 16 – 24 weeks to study the possible relationship between DNA and RNA viral load.
Methods
The amount of proviral DNA was quantified by SYBR green real-time PCR in peripheral blood mononuclear cells from a selected group of ten patients with different levels of plasmatic viremia (RNA viral load).
Results
Variable levels of proviral DNA were found without any significant correlation between proviral load and plasma HIV-1 RNA levels. Results obtained showed an increase or a rebound in viral DNA in most patients, suggesting that the absence of therapy reflects an increase and/or a persistence of cells containing viral DNA.
Conclusion
Even though plasma HIV RNA levels remain the basic parameter to monitor the intensity of viral replication, the results obtained seem to indicate that DNA levels could represent an adjunct prognostic marker in monitoring HIV-1 infected subjects.
==== Body
Background
Many papers have clearly demonstrated that HIV-1 RNA plasma viral load quantitative determination is a pivotal parameter to monitor viral replication and the effectiveness of HAART therapy [1-5]. In addition, a growing number of observations showed that measurement of HIV-1 DNA proviral load could provide crucial information on the reservoir and dynamics of HIV-1 infection [5,6] since the persistence of HIV-DNA in peripheral blood mononuclear cells (PBMC) and lymph nodes is a major drawback to eradication of infection [7,8]. Quantitative analysis of proviral DNA in HAART-treated patients showed opposite results: on one hand, the decline in DNA load seemed to indicate the long term impact and effectiveness of retroviral treatment [9-12], on the other DNA levels remained stable over several years in PI ART naïve patients [2,13].
Recent studies also indicate that viral replication persists even in individuals with prolonged suppression of plasma HIV-1 RNA levels to fewer than 50 copies/ml [5,8,14-16], confirming that "undetectable viremia" cannot be considered evidence of complete viral replication suppression. The findings of a slow and/or incomplete decay imply that current HAART regimens do not completely suppress viral replication. However the decreasing morbidity and mortality in HAART-treated HIV-1 seropositive patients and the following restoration, preservation of immunologic function and improvement in quality of life demand the ongoing use of these drugs. The new therapeutic challenge is to find new immunological or pharmacological approaches aimed at purging HIV-1 DNA proviral reservoirs [19]. Several recent studies have addressed structured treatment interruption (STI), as conceivable strategy to stimulate and enhance the immune system HIV-1 specific response to tackle viral replication in the absence of chemotherapy [17-19] even though several reports showed that only 10–20% of chronically infected patients achieved a short-term suppression of viral replication [20-22].
Since a growing number of studies involving quantification of cellular HIV-1 DNA acknowledge the importance of accurate quantification of proviral DNA in peripheral blood cells for monitoring diseases progression, we selected a small but peculiar group of patients, who decided to interrupt antiretroviral therapy, irrespective of current guidelines [23] and despite virologic failure. In particular, sequential blood samples of 10 patients under antiretroviral treatment from 1997 with two NRTIs, who refused to continue any antiviral regimen, were analyzed for 16 – 24 weeks to study the possible relationship between DNA and RNA viral load.
Methods
Patients
Ten HIV-1 infected adults under antiretroviral treatment since 1997 with two NRTIs (stavudine [D4T] and lamivudine [3TC] or zidovudine [AZT] and lamivudine [3TC] or zidovudine [AZT] and zalcitabina [DDC] or zidovudine [AZT] and didanosine [DDI]) were selected for this study. All these patients refused to continue any antiviral regimen despite an assessed virologic failure (HIV-1 RNA viral load > 50 copies/ml) and were followed-up monthly for a variable period ranging from 16 to 24 weeks up to the moment in which they agreed to begin a new therapeutic protocol. Sequential blood samples were obtained at baseline (time 0: voluntary therapy interruption) and every four weeks (time 1-time 7) and analyzed for viral load (RNA and DNA), and CD4 levels. The baseline characteristics of the patients included in the study are shown in Table 1. All the subjects were enrolled after informed consent according the Helsinki declaration of 1975.
Table 1 Baseline characteristics of HIV infected patients enrolled in the study at time of therapy suspension.
Characteristic
Gender 7 males, 3 females
Heterosexual 8
IVDUs 2
Age (mean years ± SD) 36.27 ± 8.32
CD4 count cells × 106 per l (median) 548.45 ± 63 cells/mmc.
Plasma HIV-1 RNA copies/ml (median) 3.7 × 103
HIV-1 RNA quantification
All the whole blood samples were centrifuged at 2500 rpm for 20 min and plasma was stored at -80°C until use. Plasma was analyzed for HIV-1 RNA viral using the Quantiplex HIV-RNA-3.0 assay (Chiron Corporation, Emeryville, CA, USA), according to the manufacturer's instructions. The amount of HIV RNA levels was expressed as copy number per ml of plasma and the lowest detection limit of the assay was 50 copies/ml.
DNA extraction and purification of PBMCs for HIV-1 DNA quantification
Peripheral blood mononuclear cells (PBMC) were isolated from whole blood by Ficoll-Paque gradient separation (Amersham Pharmacia). Cell pellets, corresponding to 5 × 106 PBMC were prepared and stored at -80°C. DNA was extracted and purified from each PBMC pellet by DNAeasy tissue kit (Qiagen) following the manufacturers' instructions. The pGEMBH10 HIV plasmid [8,14] was purified by Midi plasmid extraction kit (Qiagen, Hilden, Germany) following the manufacturers' instructions. Plasmid and cellular DNA concentration and purity were determined by spectrophotometric analysis at 260/280 nm.
Determination of HIV-1 proviral DNA by SYBR green real-time PCR
SYBR green real-time PCR assay was performed, as previously described [8,14] in 20 μl PCR mixture volume consisting in 2× Quantitect SYBR Green PCR Master Mix (Qiagen) containing HotStarTaq DNA polymerase, 200 nM of each oligonucleotide primer (SK431, SK462) [24] and 600 ng of DNA extracted from clinical samples (approximately the DNA content of 200.000 cells) or scalar dilution of pGEMBH10 HIV plasmid (from 105 to 10 copies). Initial activation of HotStar Taq DNA Polymerase at 95°C for 15 min; 45 cycles in four steps: 94°C for 10 s, 60°C for 30 s, 72°C for 30 s, 78°C for 3 s. At the end of amplification cycles, melting temperature analysis was carried out by a slow increase in temperature (0.1°C/s) up to 94°C. Amplification, data acquisition and analysis were carried out by a LightCycler instrument (Roche, Mannheim, Germany) using LightCycler 5.3.2 software (Roche). This software, coupled with the LightCycler instrument, determines the threshold cycle (Ct) representing the number of cycles in which the fluorescence intensity is significantly above the background fluorescence. Ct is directly proportional to log10 of the copy number of the input templates with respect to a standard curve generated in parallel. SYBR green molecules bind all double stranded DNA molecules emitting a fluorescent signal, on binding, proportional to the amplicon synthesis during the PCR reaction. This property elicited an accurate analysis of the melting temperature curve of the amplified fragments generated by real-time PCR to determine the detection and quantitation of specific products. Thus the single analysis of fluorescence was performed at 75°C by LightCycler 5.3.2 software in each cycle to rule out any non-specific interference (i.e. dimer primer). All samples from patients were run in duplicate and were also analyzed by SYBR Green real-time PCR for globin gene in a parallel run to check the equal amount in all samples determined by spectrophotometric data as described.
Statistical analysis
Statistical analysis was carried out using Student's t-test or Mann-Whitney test. Correlation was determined by Spearman's rank correlation.
Results
Longitudinal analysis of RNA plasma viral load detection by b-DNA assay
As expected, therapy interruption determined a significant increase in RNA viral load in all HIV-1 seropositive patients enrolled in the study. In particular, all patients' plasma showed a significant (Mann-Whitney test p = 0.036) increase in viral load already one month after therapy interruption (time 1) showing a median value of 1.7 × 104 (4.2 log10) in comparison to 3.7 × 103 copies/ml (3.5 log10) observed at median baseline value (time 0). Moreover, plasma viral load reached higher levels [median value of 1,1 × 105 HIV-1RNA copies/ml (5 log10)] at the end of observation period (p = 0.014) (Table 2). Hence, we assessed an increase in viral replication ranging from 0.5 log10 to more than 1 log10 at the end of observation period (p = 0.00).
Table 2 Longitudinal values of HIV RNA and DNA viral load, CD4 levels in patients on long term treatment with two NRTIs from therapy suspension (time 0) onwards.
RNA Viral Load (copies/ml) DNA Viral Load (copies/106PMBCs) CD4 cell count (x106cells/L) RNA Viral Load (copies/ml) DNA Viral Load (copies/106PMBCs) CD4 cell count (x106cells/L)
Pt 1 Pt6
Time0 6300 335 Time0 2940 625
Time1 20800 1600 324 Time1 2500 1100 625
Time2 30000 2200 324 Time2 16000 1400 529
Time3 58000 1800 340 Time3 7800 630 625
Time4 100000 2000 300 Time4 25000 620 552
Time5 120000 3800 220 Time5 24000 730 750
Pt2 Time6 17000 890 483
Time0 11000 345 Pt 7
Time1 22000 1200 345 Time0 1600 462
Time2 70000 1900 365 Time1 6800 920 450
Time3 100000 1500 256 Time2 1700 750 456
Time4 150000 1000 304 Time3 4400 910 384
Time5 148000 770 240 Time4 4000 690 380
Time6 170000 990 238 Time5 2700 400 399
Pt3 Time6 2800 300 342
Time0 3600 551 Pt8
Time1 43000 2300 448 Time0 3970 616
Time2 58100 820 266 Time1 4000 1000 616
Time3 49000 580 560 Time2 85000 660 532
Time4 58000 990 360 Time3 100000 490 361
Pt4 Time4 84000 230 420
Time0 1220 572 Time5 74000 330 456
Time1 1400 1300 570 Time6 100000 550 460
Time2 24000 1890 522 Pt9
Time3 46000 960 572 Time0 10500 680
Time4 27000 500 576 Time1 14000 1500 624
Time5 32000 700 432 Time2 16000 1000 480
Time6 31000 550 348 Time3 39000 450 588
Pt5 Time4 33000 1100 520
Time0 4900 494 Time5 145000 3100 390
Time1 93000 1500 528 Pt10
Time2 100000 1050 483 Time0 1130 578
Time3 150000 830 460 Time1 69000 612 420
Time4 250000 1000 404 Time2 66000 1200 320
Time3 100000 1600 350
Time4 210000 2800 267
Longitudinal analysis of PBMC DNA proviral load detection by quantitative real time PCR assay
In parallel experiments, we quantified proviral DNA load in PBMC isolated from patients' whole blood sequential samples at fixed times after therapy suspension. The median number of samples available for each patient was five [1 month after the therapy suspension (time 1) and then each month up to the end of observation period], ranging from two to seven. The median follow-up was 5.5 (4–7) months.
The majority of patients showed a fluctuating trend in DNA viral load. Three patients (N°1, N°9 and 10) showed an increase in DNA viral load detectable from the first through to the last available sample. Even though DNA amount reached a significantly (considered as a variation of 0.5 – 1 log10) higher value only in samples from patients N°1 and N°10 (from 3.2 log10 to 3.7 log10 and from 2.7 log10 to 3.4 log10 respectively), sequential PBMC samples obtained from patient N°9 exhibited a clear tendency to increase (and from 3.1 log10 to 3.5 log10 copy of HIV-1 DNA per 106 PBMCs respectively) redoubling the DNA content. Moreover, most of the other samples obtained from patients N°2, N°3, N°4, N°5 and N°6 showed a swinging course. After an apparent decline in proviral DNA content during the follow-up, in the latest samples a moderate increase in proviral DNA load was observed in PBMC from all patients. In contrast, a decrease of HIV-1 proviral DNA content was noticed from a baseline value of 1.3 × 103 copies per ml (3.1 log10) to 5.5 × 102 HIV-1 DNA copies per ml (2.7 log10) and from 9.2 × 102 (2.9 log10) to 3.0 × 102 (2.4 log10) in patients N°4 and 7 only.
As expected, a statistical analysis of PBMCs HIV proviral DNA content and plasma RNA viral load of all 10 patients failed to disclose any significant correlation between HIV-1 proviral DNA load and HIV-1 RNA viral load (Mann-Wittney test) confirming our previous data [14].
CD4 cell count determination
All the patients enrolled in the study showed a CD4 reduction during the follow-up. All patients, except two (Patients N°1 and N°2), interrupted therapy with a level of CD4 cells >400 cells/mmc and, as expected, showed a sharp [(N°1 and N°3 (34% reduction), N°2 (31% reduction), N°4 (39%), N°9 (42%), N°10 (53%)] decrease or a moderate decline [N°5 (18% reduction), N°6 (22% reduction) N°7 and 8 (25% reduction)] at the end of our observation period. No correlation was found (r = 0.5, p > 0.005) between the course of DNA viral load and CD4 levels, but high RNA levels were significantly associated with lower CD4 counts, demonstrating a significant inverse correlation between CD4+ cell counts and HIV-1 RNA levels (p = 0.001).
Discussion
During recent years, planned therapy interruption has been entertained in specific clinical situations even though the potential role of this choice with respect to the balance between risk of disease progression and potential benefits remains to be elucidated. Our study focused on a peculiar group of patients who voluntary opted to suspend antiretroviral therapy for a variable period of time, ranging from five to seven months, despite of virologic failure. Our follow-up ceased when patients agreed to a new therapeutic protocol.
Our study aimed to evaluate the virologic evolution of these subjects focusing on DNA proviral load course, since accurate quantification of HIV-1 DNA in peripheral blood cells is an important parameter for monitoring disease progression and predicting the clinical outcome of infection [3,27-29]. Several studies, mostly addressed to patients under different therapy protocols, have shown that the evaluation of DNA content may have important implications for understanding the virological response to combination therapy [25,26]. Even thought the plasma HIV-1 RNA load is widely considered a direct indicator of viral replication in infected individuals, the formation, stability and turnover of potentially infectious virus in the HIV-1 DNA proviral pool has important indication for the understanding of HIV pathogenesis [5,6,11]. Moreover, Vitone et al. [8] recently demonstrated that the decrease in HIV-1 DNA proviral load is inversely correlated to CD4 level in HIV-1 seropositive patients with a persistently undetectable viremia (HIV-1 RNA viral load).
Current data on course of DNA viral load during infection are inconclusive [1-9], but most studies suggest that HIV-1 DNA proviral quantification is useful to monitor the decay of the HIV reservoir towards disease remission, distinguishing "responder" from "non responder" patients [3,28].
Our results, obtained from patients, therapy-free during the virolgical follow-up, showed a viral rebound, one month after therapy suspension, assessed by plasma RNA values in all patients. The analysis of HIV-1 DNA proviral content displayed a clear increase from the baseline value in three patients, confirming that an active viral replication results in elevated viremia (HIV-1 RNA load) and in an increased number of cells containing viral DNA [27,28]. Also, patients who showed an apparent decrease in DNA copy number during the first step of our follow-up, came to present a rebound of DNA in PBMCs at the end of observation period. These observations might suggest that previous therapy controlled the amount of viral DNA only for a limited period of time and a likely viral rebound, as assessed by an increase in DNA amount, was observed only some months later.
Finally, in contrast with other patients, two subjects showed a clear HIV-1 DNA proviral decrease over time, in the absence of therapy and a steady HIV-1 RNA viral load detectable in plasma samples. In both cases a HIV-1 DNA proviral decline due to a long lasting effect of therapy could be ruled out, since both patients showed high levels of viral replication by increasing value of HIV-1 RNA viral load over time. In an attempt to explain the course of HIV-1 DNA proviral in these subjects, we had to take into consideration that our assay, a SYBR green based real time PCR measures both integrated and unintegrated HIV-1 DNA form on PBMCs. There is evidence that only a fraction of integrated and unintegrated HIV-1 DNA is replication competent [25]. Hence, it is possible that most of the HIV-1 DNA, displayed in our two patients, might be mainly represented by integrated DNA fully capable of initiating HIV replication. Our data are confined to results related to proviral DNA in PBMCs, even if we must consider that viral load is also sustained by lymph node trapped CD4 T cells and other non circulating elements [6] that preserve replication competent virus for long periods. In the absence of therapy, a large number of HIV-1 DNA proviral copies might replicate, as assessed by the HIV-1 RNA viral load increase, leading to a relative decline of cellular DNA. In addition, we cannot exclude a further increase in DNA content in a longer follow-up.
Despite contrasting reports on the meaning of DNA proviral content in HIV-1 seropositive patients [2,5,9-12], our data obtained on closely controlled patients, emphasize the interest of studying DNA proviral content in HIV-1 infected patients. Even though it is impossible to define a proviral DNA threshold for use in clinical practice, several data showed that patients with high proviral DNA levels are more likely to experience virological failure than those with lower proviral DNA loads [11]. Moreover the proviral load probably reflects individual parameters because host genetic factors and response to treatment probably are involved in the constituting the pool of infected cells [30,31]. Although RNA viral load provides important information on viral replication, HIV-1 DNA proviral load can be considered an additional marker to provide crucial information, not only during the follow-up of patients under therapy but also for individuals included in structured therapy interruptions protocols. Data obtained from our patients, who were not part of antiretroviral protocols [23], yield important information on the persisting timing of DNA in PBMCs.
Conclusion
Only careful evaluation of virological and immunological markers is necessary to fully characterize the course of HIV-1 infection and to provide a more complete laboratory-based assessment of disease progression. However, the availability of a new standardized assay such as DNA proviral load will be important to assess the true extent of virological suppression in patients with non-quantifiable plasma viral loads and to verify the efficacy of new immune-based therapies aimed at purging HIV-1 DNA reservoirs. Although the biological meaning of DNA proviral load in PBMCs is not yet clear, several studies [2,3,6,10] suggest that HIV-1 cellular DNA load may be an indicator of spread of infection whereas the plasma RNA load is indicates active infection [2]. However the qualitative and quantitative evaluation of both plasma HIV RNA genome and HIV-1 proviral DNA might prove crucial to understanding the course of HIV-1 infection.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MCR and DG conceived and designed the study. FS and PV developed the HIV-1 DNA real time and performed all the experimental work. LS and FG provided blood samples and clinical information on the patients enrolled in this study. MCR drafted the manuscript and DG reviewed it. All authors contributed to the final version of manuscript, read and approved it.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by the "AIDS projects" of the Italian Ministry of Health, funds for selected research topics of the University of Bologna and MURST 60%.
We thank Ms Anne Collins for editing the manuscript.
==== Refs
Brostrom C Visco-Comandini U Yun Z Sonnerborg A Longitudinal quantification of human immunodeficiency virus type 1 DNA and RNA in long-term nonprogressors J Infect Dis 1999 179 1542 1548 10228079 10.1086/314757
Kostrikis LG Touloumi G Karanicolas R Pantazis N Anastassopoulou C Karafoulidou A Goedert JJ Hatzakis A Multicenter Hemophilia Cohort Study Group Quantitation of human immunodeficiency virus type 1 DNA forms with the second template switch in peripheral blood cells predicts disease progression independently of plasma RNA load J Virol 2002 76 10099 10108 12239284 10.1128/JVI.76.20.10099-10108.2002
McDermott JL Giri AA Martini I Bono M Giacomini M Campelli A Tagliaferro L Cara A Varnier OE Level of human immunodeficiency virus DNA in peripheral blood mononuclear cells correlates with efficacy of antiretroviral therapy J Clin Microbiol 1999 37 2361 2365 10364617
Riva E Antonelli G Scagnolari C Pistello M Capobianchi MR Monforte A Pezzotti P Dianzani F I.CO.N.A. Study Group Human immunodeficiency virus (HIV) DNA load and level of immunosuppression in treatment-naïve HIV-1-infected patients J Infect Dis 2003 187 1826 1828 12751043 10.1086/375251
Ramratnam B Mittler JE Zhang L Boden D Hurley A Fang F Macken CA Perelson AS Markowitz M Ho DD The decay of the latent reservoir of replication-competent HIV-1 is inversely correlated with the extent of residual viral replication during prolonged anti-retroviral therapy Nat Med 2000 6 82 85 10613829 10.1038/71577
Nickle DC Shriner D Mittler JE Frenkel LM Mullins JI Importance and detection of virus reservoirs and compartments of HIV infection Curr Opin Microbiol 2003 6 410 416 12941414 10.1016/S1369-5274(03)00096-1
Brun-Vezinet F Boucher C Loveday C Descamps D Fauveau V Izopet J Jeffries D Kaye S Krzyanowski C Nunn A Schuurman R Seigneurin JM Tamalet C Tedder R Weber J Weverling GJ HIV-1 viral load, phenotype, and resistance in a subset of drug-naive participants from the Delta trial. The National Virology Groups. Delta Virology Working Group and Coordinating Committee Lancet 1997 350 983 90 9329513 10.1016/S0140-6736(97)03380-1
Vitone F Gibellini D Schiavone P Re MC Quantitative DNA proviral detection in HIV-1 patients treated with antiretroviral therapy J Clin Virol 2005 15911440
Chun TW Stuyver L Mizell SB Ehler LA Mican JA Baseler M Lloyd AL Nowak MA Fauci AS Presence of an inducible HIV-1 latent reservoir during highly active antiretroviral therapy Proc Natl Acad Sci U S A 1997 94 13193 13197 9371822 10.1073/pnas.94.24.13193
Finzi D Hermankova M Pierson T Carruth LM Buck C Chaisson RE Quinn TC Chadwick K Margolick J Brookmeyer R Gallant J Markowitz M Ho DD Richman DD Siliciano RF Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy Science 1997 278 1295 300 9360927 10.1126/science.278.5341.1295
Pellegrin I Caumont A Garrigue I Merel P Schrive MH Fleury H Dupon M Pellegrin JL Ragnaud JM Predictive value of provirus load and DNA human immunodeficiency virus genotype for successful abacavir-based simplified therapy J Infect Dis 2003 187 38 46 12508144 10.1086/345860
Wong JK Hezareh M Gunthard HF Havlir DV Ignacio CC Spina CA Richman DD Recovery of replication-competent HIV despite prolonged suppression of plasma viremia Science 1997 278 1291 1295 9360926 10.1126/science.278.5341.1291
Hockett RD JrSaag MS Kilby JM Sfakianos G Wakefield TB Bucy RP Stability in the HIV vDNA pool in peripheral CD4+ T cells of untreated patients by single tube quantitative PCR J Virol Methods 2000 87 1 12 10856747 10.1016/S0166-0934(00)00139-7
Gibellini D Vitone F Schiavone P Ponti C La Placa M Re MC Quantitative detection of human immunodeficiency virus type 1 (HIV-1) proviral DNA in peripheral blood mononuclear cells by SYBR green real-time PCR technique J Clin Virol 2004 29 282 9 15018857 10.1016/S1386-6532(03)00169-0
Furtado MR Callaway DS Phair JP Kunstman KJ Stanton JL Macken CA Perelson AS Wolinsky SM Persistence of HIV-1 transcription in peripheral-blood mononuclear cells in patients receiving potent antiretroviral therapy N Engl J Med 1999 340 1614 22 10341273 10.1056/NEJM199905273402102
Zhang L Ramratnam B Tenner-Racz K He Y Vesanen M Lewin S Talal A Racz P Perelson AS Korber BT Markowitz M Ho DD Quantifying residual HIV-1 replication in patients receiving combination antiretroviral therapy N Engl J Med 1999 340 1605 1613 10341272 10.1056/NEJM199905273402101
Florence E Garcia F Plana M Fumero E Castro P Lopez A Cruceta A Gil C Arnedo M Miro JM Pumarola T Gallart T Gatell JM Long-term clinical follow-up, without antiretroviral therapy, of patients with chronic HIV-1 infection with good virological response to structured treatment interruption Clin Infect Dis 2004 39 569 574 15356824 10.1086/422720
Yerly S Gunthard HF Fagard C Joos B Perneger TV Hirschel B Perrin L Swiss HIV Cohort Study Proviral HIV-DNA predicts viral rebound and viral setpoint after structured treatment interruptions AIDS 2004 18 1951 1953 15353981 10.1097/00002030-200409240-00011
Gallo RC New concepts for future control of HIV derived from studies of pathogenesis Int J Antimicrob Agents 2000 16 421 423 11118851 10.1016/S0924-8579(00)00297-1
Oxenius A Hirschel B Structured treatment interruptions in HIV infection: benefit or disappointment? Expert Rev Anti Infect Ther 2003 1 129 139 15482106 10.1586/14787210.1.1.129
Carcelain G Tubiana R Samri A Calvez V Delaugerre C Agut H Katlama C Autran B Transient mobilization of human immunodeficiency virus (HIV)-specific CD4 T-helper cells fails to control virus rebounds during intermittent antiretroviral therapy in chronic HIV type 1 infection J Virol 2001 75 234 241 11119593 10.1128/JVI.75.1.234-241.2001
Ortiz GM Wellons M Brancato J Vo HT Zinn RL Clarkson DE Van Loon K Bonhoeffer S Miralles GD Montefiori D Bartlett JA Nixon DF Structured antiretroviral treatment interruptions in chronically HIV-1-infected subjects Proc Natl Acad Sci U S A 2001 98 13288 13293 11687611 10.1073/pnas.221452198
Guidelines for the Use of Antiretroviral Agents in HIV-1-infected adults and adolescents October 29, 2004
Christopherson C Kidane Y Conway B Krowka J Sheppard H Kwok S PCR-Based assay to quantify human immunodeficiency virus type 1 DNA in peripheral blood mononuclear cells J Clin Microbiol 2000 38 630 634 10655358
Hatzakis AE Touloumi G Pantazis N Anastassopoulou CG Katsarou O Karafoulidou A Goedert JJ Kostrikis LG Cellular HIV-1 DNA load predicts HIV-RNA rebound and the outcome of highly active antiretroviral therapy AIDS 2004 18 2261 2267 15577538 10.1097/00002030-200411190-00006
Tierney C Lathey JL Christopherson C Bettendorf DM D'Aquila RT Hammer SM Katzenstein DA Prognostic value of baseline human immunodeficiency virus type 1 DNA measurement for disease progression in patients receiving nucleoside therapy J Infect Dis 2003 187 144 8 12508159 10.1086/345870
Strain MC Little SJ Daar ES Havlir DV Gunthard HF Lam RY Daly OA Nguyen J Ignacio CC Spina CA Richman DD Wong JK Effect of Treatment, during Primary Infection, on Establishment and Clearance of Cellular Reservoirs of HIV-1 J Infect Dis 2005 191 1410 1418 15809898 10.1086/428777
Kabamba-Mukadi B Henrivaux P Ruelle J Delferriere N Bodeus M Goubau P Human immunodeficiency virus type 1 (HIV-1) proviral DNA load in purified CD4+ cells by LightCycler real-time PCR BMC Infect Dis 2005 21 15 15780144 10.1186/1471-2334-5-15
Wong JK Hezareh M Gunthard HF Havlir DV Ignacio CC Spina CA Richman DD Recovery of replication-competent HIV despite prolonged suppression of plasma viremia Science 1997 278 1291 1295 9360926 10.1126/science.278.5341.1291
Berger EA Murphy PM Farber JM Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease Annu Rev Immunol 1999 17 657 700 10358771 10.1146/annurev.immunol.17.1.657
O'Brien TR McDermott DH Ioannidis JP Carrington M Murphy PM Havlir DV Richman DD Effect of chemokine receptor gene polymorphisms on the response to potent antiretroviral therapy AIDS 2000 14 821 6 10839590 10.1097/00002030-200005050-00008
|
16156892
|
PMC1236926
|
CC BY
|
2021-01-04 16:28:14
|
no
|
BMC Infect Dis. 2005 Sep 12; 5:69
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-69
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-701616228910.1186/1471-2334-5-70Research ArticleMortality following Campylobacter infection: a registry-based linkage study Ternhag Anders [email protected]örner Anna [email protected] Åke [email protected] Johan [email protected] Karl [email protected] Department of Epidemiology, Swedish Institute for Infectious Disease Control, Solna, Sweden2 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden3 Division of Mathematical Statistics, Stockholm University, Sweden2005 14 9 2005 5 70 70 21 1 2005 14 9 2005 Copyright © 2005 Ternhag et al; licensee BioMed Central Ltd.2005Ternhag 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
Campylobacteriosis is one of the most commonly identified causes of bacterial diarrheal disease and a common cause of gastroenteritis in travellers from developed nations. Despite the widespread occurrence, there is little information on Campylobacter mortality.
Methods
Mortality among a cohort of Campylobacter cases were compared with the general population 0–1, 1–3, 3–12 and more than 12 month after the onset of the illness. The cases were sub-grouped according to if they had been infected domestically or abroad.
Results
The standardized mortality ratio for cases infected domestically was 2.9 (95% CI: 1.9–4.0) within the first month following the illness. The risk then gradually diminished and approached 1.0 after one year or more have passed since the illness. This initial excess risk was not attributable to any particular age group (such as the oldest).
In contrast, for those infected abroad, a lower standardized mortality ratio 0.3 (95% CI: 0.04–0.8) was shown for the first month after diagnosis compared to what would be expected in the general population.
Conclusion
Infection with Campylobacter is associated with an increased short-term risk of death among those who were infected domestically. On the contrary, for those infected abroad a lower than expected risk of death was evident. We suggest that the explanation behind this is a "healthy traveler effect" among imported cases, and effects of a more frail than average population among domestic cases.
==== Body
Background
Campylobacteriosis is a zoonotic disease observed in most parts of the world. The disease is caused by Campylobacter jejuni, or less commonly Campylobacter coli. It is estimated to cause 5–14% of diarrhea, worldwide [1], and causes a great amount of hospital bed days in all countries. In the United States, campylobacteriosis account for approximately 17% of hospitalizations due to food borne infections [2], and in England and Wales Campylobacter is by far the most important food-borne pathogen when it comes to GP consultations, hospital admissions and hospital bed days [3]. In Sweden, campylobacteriosis is the most common bacterial food-borne disease. Since 1997, between 6,900 and 8,600 annual cases (78–96 per 100,000) have been reported to the Swedish Institute for Infectious Disease Control (SMI), of which approximately 55% were imported [4].
Apart from diarrhea, many patients experience abdominal pain, often accompanied by fever, malaise and headache. The diarrhea can range from mild to profuse, causing dehydration. There are also some known although rare complications to Campylobacter infection; Guillain-Barré syndrome is thought to occur after 1 in 1,000 cases of C. jejuni infection [5,6]. This involves autoantibodies against distinct surface polysaccharides that are similar to host ganglioside structures [7]. Reactive arthritis and Reiter's syndrome (asymmetric arthritis, urethritis and ophthalmitis) are other extra-intestinal manifestations caused by Campylobacter, which affects approximately 1% of cases [8].
Although Campylobacter infection is both very common and can sometimes result in severe complications, less is known about the mortality attributable to Campylobacter infection. In this study we therefore via registry linkage estimated the mortality in a cohort of Campylobacter cases in Sweden. The primary objective was to investigate if cases diagnosed with Campylobacter have a mortality rate from all causes that differs from the general population. If so, the second objective was to see if this all cause mortality rate among campylobacterosis cases mirrors Campylobacter related mortality rates. We would in this case expect any elevated mortality rate in the Campylobacter cohort to be highest close to infection and then gradually diminish with time and approach the general mortality rate in the population.
Methods
Sweden has a dual reporting system for notifiable infectious diseases. This means that both the doctor having seen the patient and the laboratory having isolated the pathogen are required to file a report to the SMI. Using a unique personal identification number, issued to all Swedish residents, and used in all contacts with the health care, reports from the two sources are merged into case files.
For this study, we identified all persons reported with campylobacteriosis to the SMI between January 1, 1997 and December 31, 2003. For each case we extracted date of birth, sex, country of infection, date of disease onset, date of diagnosis and reporting date from the database. If the date for disease onset or date of diagnosis was missing we used the median time from patients who have all three dates to back-calculate a date for disease onset. These cases were then via the personal identification number linked to a database that contain information from the National Tax Board on every deceased person in Sweden and their date of death.
We then calculated follow-up time per age group, from the date of disease onset until either an event took place (death) or until November 1, 2004. The following time strata were used; 0–1 month, 1–3 months, 3–12 months and more than 12 months after the onset of campylobacteriosis. The first time strata was constructed to be long enough so that acute effects of even prolonged illness was not missed. The other cut-off was chosen so that effects between strata could be evident (for example a declining risk with time) and that each time strata contained a sufficient amount of person-time for robust estimations. From Statistics Sweden we obtained sex- and age-specific death rates, which were used to obtain the expected number of deaths. Standardized mortality ratios (SMRs) were calculated by dividing the observed number of deaths in our cohort with the expected number of deaths. If the result was >1.0, then the mortality in the Campylobacter cohort was higher than in the general population, and conversely, if the ratio was <1.0 then the mortality among our cases was lower than in the general population. The SMR is a weighted average of all age groups in the Campylobacter cohort. This indirect standardization was done with 5-year age strata (e.g. 0–4, 5–9, etc.) but is presented here either as SMR, or grouped into broader categories e.g. 0–14 years, 14–64 years and +65 years.
With the assumption that those who had acquired their infection abroad and those who had contracted it at home were two fundamentally different groups – the former probably healthier than the latter ("healthy traveler effect") – we divided our Campylobacter cohort into two groups, imported and domestic cases [9].
Exact confidence intervals and tests were calculated assuming that the number of deaths in each stratum was Poisson distributed. Within each time stratum the observed SMRs for each age group were compared using an exact test. The observed number of cases was assumed to be multinomially distributed over the age classes with probabilities proportional to the expected values. All calculations were performed using SAS statistical software, V. 8.2.
Results
During the study period 1997–2003, 48,025 unique individuals were notified with Campylobacter infection, of whom 16,710 were infected in Sweden, and 28,930 were infected abroad. For the remaining 2,385 cases either the personal identification number was incomplete, or the country of infection was unknown.
For cases reported with Campylobacter and infected in Sweden, the SMR was higher compared to the general Swedish population (Table 1). The risk was highest within the first month after onset of disease (SMR 2.9; 95% CI 1.9–4.0). After that time period the risks gradually decreased and after one year had passed after the debut of illness, the risk was the same as for the general population (SMR 1.0; 95% CI 0.9–1.1)
Table 1 Standardized mortality ratios (SMR) during the period 1997–2003 among 45,640 Swedish reported cases of campylobacteriosis, infected in Sweden (n = 16,710) and abroad (n = 28,930)
Infected in Sweden Infected abroad
Time after infection (months) Obs* (n = 563) Exp* SMR* (95% CI*) Obs (n = 222) Exp SMR (95% CI)
0–1 30 10.4 2.9 (1.9–4.0) 2 6.8 0.3 (0.04–0.8)
1–3 31 20.7 1.5 (1.0–2.1) 10 13.7 0.7 (0.4–1.3)
3–12 123 93.9 1.3 (1.1–1.6) 28 63.3 0.4 (0.3–0.6)
>12 379 388.6 1.0 (0.9–1.1) 182 330.6 0.6 (0.5–0.6)
* Obs, observed number of deaths; Exp, expected number of deaths; SMR, standardized mortality ratio; CI, confidence interval
To see if this early excess mortality risk was attributable to any particular age group within the group of domestically infected individuals, we divided the cohort into three age groups (0–14 years, 15–64 years and 65 years and above). Table 2 shows the mortality ratios for these separate age groups. No significant differences between the groups were found at any of the time periods following the illness (the lowest p-value is 0.20). Persons infected with Campylobacter abroad, had lower SMR compared to the general population (Table 1). We observed less number of deaths than expected in every time stratum after the debut of illness. Within the first month after the disease SMR was 0.3 (95% CI 0.04–0.8), and a significantly lower risk compared to the reference population was evident for every time period except for the second time stratum (1–3 months after infection) where the SMR was 0.7 (95% CI: 0.4–1.3).
Table 2 Standardized mortality ratios (SMR) in age groups 0–14, 15–64 and 65+ years among 16,710 domestically infected Campylobacter cases in Sweden 1997–2003
Time after infection Age group (years) Obs* Exp* SMR* (95% CI*)
0–1 months 0–14 1 0.1 8.9 (0.2–32.9)
15–64 4 2.0 2.0 (0.5–4.3)
65+ 25 8.3 3.0 (2.0–4.3)
1–3 months 0–14 0 0.2 0 (0–13.5)
15–64 7 4.1 1.7 (0.7–3.2)
65+ 24 16.4 1.5 (0.9–2.1)
3–12 months 0–14 3 1.0 3.1 (0.6–7.6)
15–64 26 18.5 1.4 (0.9–2.0)
65+ 94 74.4 1.3 (1.0–1.5)
>12 months 0–14 0 2.3 0 (0–1.3)
15–64 72 75.5 1.0 (0.7–1.2)
65+ 307 310.7 1.0 (0.9–1.1)
* Obs, observed number of deaths; Exp, expected number of deaths; SMR, standardized mortality ratio; CI, confidence interval
Discussion
The SMR was almost three times higher for persons infected with Campylobacter in Sweden within the first months after disease onset. These risks diminished with time, and after a year or more have passed the SMR almost equaled that for the general population. For the imported cases of Campylobacter we observed lower SMRs throughout every time point in the study period.
Case fatality rate (CFR) is widely used and perhaps more intuitive than standardized mortality ratio when disease specific mortality is estimated. Up until November 1, 2004, 563 of the cases infected in Sweden, and 222 of the cases infected abroad had died. The CFR within 30 days was 0.19% (95% CI 0.13–0.27%) for those infected in Sweden, and 0.008% (95% CI 0.0008–0.03%) for those infected abroad. However, we do not favour the use of CFR because information on competing causes of mortality is missing, along with age effects.
The different SMRs depending on whether a case was infected at home or abroad must be analyzed in a context of how a case is diagnosed and reported. For gastrointestinal diseases there are problems with under-reporting and biases regarding which cases that eventually turns up in the national routine surveillance.
The problems with under-reporting affect every step in the chain: an ill person must seek health care, the doctor needs to take a stool sample, the laboratory must be capable of accurately detecting the pathogen, and if the laboratory find the pathogen the doctor needs to send in a notification. It is perhaps in the first of these steps that national surveillance misses most of the cases. Studies in both United Kingdom and United States have tried to measure the proportion reported to national surveillance from those who have symptoms of infectious intestinal disease in the population. For an agent like Campylobacter that typically causes non-bloody diarrhea a multiplier in the range of 7.6–38 has been estimated [10,11].
All persons who get infected with Campylobacter do not have the same chance of becoming notified as a Campylobacter case. There are three different groups that are more likely than others to be reported [12]. Firstly, people who have a severe disease or profound symptoms are more likely to see a doctor, compared to others. Secondly, persons with a history of recent travel prior to the onset of symptoms are also more likely to seek health care and become diagnosed and reported. Thirdly, persons with pre-existing illnesses of certain magnitude may be over-represented in the national surveillance statistics.
These biases in disease reporting will have an impact on our findings and can help us in the interpretation of the results. The general under-reporting will affect milder cases in particular. Therefore, the persons in our cohort infected in Sweden were most likely the more severe cases in the population. On the other hand, travellers who develop symptoms of gastro-enteritis are more likely on the average to have a physical exam when they have returned home and they do not necessary need to have a severe disease. The finding of a standardized mortality ratio equal to 1.0 after one year had passed after the infection suggest that the mortality of this cohort infected domestically is comparable to the standard Swedish population after one year. This does not necessarily imply that the cohort is comparable to the Swedish population also at the earlier time points. The risk profile may have changed due to frailty. The statistical term frailty implies that all individuals do not have the same risk profile and frailer individuals may succumb earlier and therefore the risk profile of a cohort changes over time [13].
It is difficult to find information on Campylobacter associated mortality. Some have estimated the case fatality rate to be in range of 1–3/10000 cases [14]. Others that have used an approach more like ours found that 1.2% (16180 cases and 190 deaths) where deceased within one year from diagnosis [15]. Unfortunately, no study has compared mortality risks among travellers and non-travellers for campylobacteriosis.
Our findings suggest that the individuals reported into the surveillance system are a highly selected. Individuals with a domestically contacted infection may have a different, more severe risk profile, compared to individuals in the cohort with an imported infection. The reason for these differences is not characteristics of the infection itself but reflects differences in how these individuals are selected into the surveillance system. Any registry-based estimation of mortality among cases of campylobacteriosis must therefore take this into account and stratify cases according to the place of infection.
Conclusion
In conclusion, we found in our study cohort two groups of persons with opposite mortality risks within the first month after a Campylobacter infection. Persons who had acquired their infection in Sweden had almost a three times higher than expected risk to die. The larger group of persons, having picked up their infection abroad, had, on the contrary, a lower than expected risk to die, likely due to a "healthy traveler effect". In estimations of mortality or burden of disease regarding campylobacteriosis one should be aware that travellers differ in many characteristics compared to non-travellers. In countries where imported cases are in majority, these two groups must be handled separately to avoid any misinterpretation of the true burden of disease.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
ATe assembled and analyzed the data and drafted the manuscript. ATo performed the statistical analyses and revised the article. AS contributed to statistical analysis and critically revised the article. KE and JG participated in the study design and critically revised the article. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Coker AO Isokpehi RD Thomas BN Amisu KO Obi CL Human campylobacteriosis in developing countries Emerg Infect Dis 2002 8 237 44 11927019
Mead PS Slutsker L Dietz V McCaig LF Bresee JS Shapiro C Griffin PM Tauxe RV Food-related illness and death in the United States Emerg Infect Dis 1999 5 607 625 10511517
CC Tam O'Brien SJ Adak GK Meakins SM Frost JA Campylobacter coli – an important foodborne pathogen J Infect 2003 47 28 32 12850159 10.1016/S0163-4453(03)00042-2
Ekdahl K Andersson Y Regional risks and seasonality in travel-associated campylobacteriosis BMC Infect Dis 2004 4 54 15569393 10.1186/1471-2334-4-54
Crushell E Harty S Sharif F Bourke B Enteric Campylobacter: Purging Its Secrets? Pediatr Res 2004 55 3 12 14605259 10.1203/01.PDR.0000099794.06260.71
McCarthy N Giesecke J Incidence of Guillain-Barré syndrome following infection with Campylobacter jejuni Am J Epidemiol 2001 153 610 4 11257070 10.1093/aje/153.6.610
Bereswill S Kist M Recent developments in Campylobacter pathogenesis Curr Opin Infect Dis 2003 16 487 491 14502003 10.1097/00001432-200310000-00017
Altekruse SF Stern NJ Fields PI Swerdlow DL Campylobacter jejuni – An emerging foodborne pathogen Emerg Infect Dis 1999 5 28 35 10081669
Kelman CW Kortt MA Becker NG Li Z Mathews JD Guest CS Holman CDJ Deep vein thrombosis and air travel: record linkage study BMJ 2003 327 1072 6 14604926 10.1136/bmj.327.7423.1072
Wheeler JG Sethi D Cowden JM Wall PG Rodrigues LC Tompkins DS Hudson MJ Roderick PJ Study of infectious intestinal disease in England: rates in the community, presenting to general practice, and reported to national surveillance BMJ 1999 318 1046 50 10205103
Voetsch AC Van Gilder TJ Angulo FJ Farley MM Shallow S Marcus R Cieslak PR Deneen VC Tauxe RV FoodNet Estimate of the Burden of Illness Caused by Nontyphoidal Salmonella Infections in the United States Clin Infect Dis 2004 38 S127 34 15095181 10.1086/381578
Tam CC Rodrigues LC O'Brien SJ The study of infectious intestinal disease in England: what risk factors for presentation to general practice tell us about potential for selection bias in case-control studies of reported cases of diarrhoea Int J Epidemiol 2003 32 99 105 12690019 10.1093/ije/dyg007
Hougaard P Frailty Armitage P, Colton T Encyclopedia of biostatistics 1998 Hoboken NJ: Wiley 1572 6
Havelaar AH de Wit MAS van Koningsveld R van Kempen E Health burden in the Netherlands due to infection with thermophilic Campylobacter spp Epidemiol Infect 2000 125 505 522 11218201 10.1017/S0950268800004933
Helms M Vastrup P Gerner-Smidt P Molbak K Short and long term mortality associated with foodborne bacterial gastrointestinal infections: registry based study BMJ 2003 326 357 360 12586666 10.1136/bmj.326.7385.357
|
16162289
|
PMC1236927
|
CC BY
|
2021-01-04 16:28:13
|
no
|
BMC Infect Dis. 2005 Sep 14; 5:70
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-70
|
oa_comm
|
==== Front
BMC Med GenetBMC Medical Genetics1471-2350BioMed Central London 1471-2350-6-311614303910.1186/1471-2350-6-31Research ArticleNo evidence for the association of DRD4 with ADHD in a Taiwanese population within-family study Brookes Keeley-Joanne [email protected] Xiaohui [email protected] Chih-Ken [email protected] Yu-Shu [email protected] Yu-Yu [email protected] Philip [email protected] MRC Social Genetic Developmental Psychiatry Centre, Institute of Psychiatry, London UK2 Department of Psychiatry, Chang Gung Memorial Hospital, Taiwan2005 5 9 2005 6 31 31 14 1 2005 5 9 2005 Copyright © 2005 Brookes 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
Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent and highly heritable childhood disorder. The dopamine D4 receptor (DRD4) gene has shown a genetic association with ADHD in Caucasian populations with meta-analysis indicating a small but significant effect across datasets. It remains uncertain whether this association can be generalised to non-Caucasian ethnic groups. Here we investigate two markers within the DRD4 gene in a Taiwanese population, the exon 3 variable number tandem repeat (VNTR) and a 5' 120 base-pair duplication.
Methods
Within-family transmission disequilibrium tests of association of the 5' 120 base-pair duplication, and exon 3 VNTR in a Taiwanese population.
Results
No evidence of association of ADHD with either polymorphism in this population was observed.
Conclusion
The DRD4 gene markers investigated were not found to be associated with ADHD in this Taiwanese sample. Further work in Taiwanese and other Asian populations will therefore be required to establish whether the reports of association of DRD4 genetic variants in Caucasian samples can be generalised to Asian populations.
==== Body
Background
Attention Deficit Hyperactivity Disorder (ADHD) is one of the most prevalent and heritable childhood behavioural disorders. Progress in identifying some of the genes involved in ADHD susceptibility has been relatively fruitful over the past decade by screening genetic variants that lie within or close to genes that regulate neurotransmitter systems, particularly dopamine pathway genes [1]. One of the more consistent findings is the association with the 7-repeat allele of a 48-base pair variable number tandem repeat (VNTR) in exon 3 of D4 receptor gene (DRD4) although there are a number of negative reports and discrepancies between case control and within family studies [2-22]. Meta-analyses of published and unpublished data indicate a small but significant association across datasets with no evidence of heterogeneity [23,24].
Other genetic variants in the DRD4 5'-regulatory region have also been reported to be associated with ADHD. Of particular interest is a 120-bp duplication that has been associated with ADHD in two studies [25,26] although negative studies have also been reported [27-29,20]. Recently our group investigated the functional significance of this repeat marker using in vitro reporter gene assays and found that the long allele conferred lower transcriptional activity than the shorter alleles in four different mammalian cell lines [30].
Both the 120 bp duplication [10,31] and the exon 3 VNTR [32-35] have also been associated with novelty seeking in a few studies. Although the association with novelty seeking is not as consistent as the association with ADHD and was not significant in a meta-analysis of available data [36], these reports remain of potential interest due the clinical association of ADHD with risk taking and stimulus seeking behaviours [37].
Here we set out to replicate these findings using a Taiwanese sample of 216 ADHD probands that had previously shown association to the dopamine transporter gene [38] (Brookes et al., in review). Grady and colleagues [19] suggest from their detailed analysis of sequence data across DRD4 that in Caucasian populations the 7-repeat allele is a relatively recent mutation that represents an independent clad and has been subject to positive selection. Their data suggests that in Asian populations the 2-repeat allele is a derivative of the Caucasian 7-repeat allele and we might therefore expect to see association between the 2-repeat allele and ADHD in Taiwanese populations [35,44]. This hypothesis was not supported by Qian et al, [17,18] who did observed a case-control association with long repeat alleles (4–6-repeat alleles; p < 0.05) in a Han Chinese population. This association, however, was not supported using within family tests of association and the association was not specific to any single allele. More recent data from Leung and colleagues does however support this hypothesis by observing a significant increase in prevalence of the 2-repeat allele in a Han Chinese ADHD sample in comparison to a control sample (p = 0.015) [45].
Methods
216 ADHD probands between the ages of 5–15 years and available parents were recruited into the study from the Child Psychiatric Clinics in the Chang Gung Memorial Hospital in Taipei area, Taiwan. A total of 192 (83.6%) were males. IQ was 50–69 in 13%, 70–89 in 44%, 90–119 in 40% and >120 in 1%. The diagnosis of ADHD was made according to DSM-IV criteria following completion of a standard maternal interview (Kiddie-SADS) [39] and completion of parent and teacher Conner's revised rating scales [40]. In all 78% had the combined subtype and 22% the inattentive subtype of ADHD.
Genotyping was carried out using standard PCR methods and analysed on 2% agarose. Detection rate of DRD4 genotypes was 87.5% for the VNTR and 96% for the 5' 120 bp duplication. Both markers were in Hardy-Weinberg Equilibrium, and no Mendelian errors were observed. The family genotypes were analysed by single marker transmission disequilibrium test (TDT), and haplotype-based haplotype relative risk test (HHRR) run in UNPHASED . UNPHASED was also used to calculate haplotype associations for phase-certain haplotypes (TDTPHASE) and for uncertain haplotypes (HHRR) in addition linkage disequilibrium between the markers.
Results
The TDT data revealed that neither marker investigated is associated with ADHD in this sample, either individually, or when combined together into a haplotype (Tables 1 and 2). The Global D' value between the two markers is low (0.2) suggesting that the two markers segregate independently from each other in this population. Refining the sample by removing those with IQ less than 70, did not improve the significance of the finding for either the exon III VNTR (TDT p = 0.55; HHRR p = 0.55), the 5'120 bp repeat (TDT p = 0.71; HHRR p = 0.73) or the haplotype of the two markers (TDT p = 0.45; HHRR p = 0.45). Subtype specific tests of association for the combined and inattentive subtypes analysed separately were non-significant (data not shown).
Discussion
In summary we did not find the DRD4 markers to be associated with ADHD in a Taiwanese sample that has previously shown association with the dopamine transporter gene. As reported in previous studies in Asian populations the VNTR 7-repeat allele was absent and we also failed to replicate the previous reported associations in a Chinese population with 4-repeat and 6-repeat alleles [17,18] and in a Taiwanese population with the 2-repeat allele [45]. There have been no previously reported studies of the 120-bp repeat polymorphism and ADHD in Asian populations. Differences in association may also be due to differential diagnosis because of difference in cultures between western and eastern civilisations [48]. However the previously reported association with the dopamine transporter gene suggests that the clinical phenotype in this sample is comparable with samples ascertained in Europe, the United States and South America.
There are several possible explanations for the observed findings. First it is entirely possible that increased risk to ADHD associated with either of the markers studied may be absent in Asian populations. The association with the 7-repeat allele reported in Caucasian populations may be dependent on the presence of the 7-repeat allele itself and therefore absent from any population with low frequency of this allele. Critical sequences that increase risk for ADHD may therefore be absent in this population. The association with the 120-bp duplication has yet to be established and these data lend no further support to this potential finding. Second, our sample lacked power to detect very small main effects with less than 80% power at alpha = 0.05 for odds ratios less than 1.5 for associations with either the 2-repeat or 4-repeat alleles. Furthermore, it is feasible that environmental risk factors might interact with genetic risk factors reducing or abolishing main effects from genotype alone (e.g. [41,42]). In this case we would have little chance of detecting such associations unless we also had measured environmental risks. Third, previous studies using within family tests of association have failed to find evidence for association of the VNTR polymorphism with ADHD, whereas case-control designs have been more positive (discussed in [1]). The work of Holmes et al [8] using a collaborative dataset suggests that although there was no preferential transmission of the 7-repeat allele, there was a significant TDT association in the sub-group that had co-morbid conduct disorder. This is consistent with a report that parent-proband trios samples have a less severe clinical phenotype with lower levels of ADHD and co-morbid symptoms [43]. The association may therefore be present in co-morbid groups or groups which have certain aspects of the ADHD phenotype such as novelty seeking, cognitive impulsiveness [22], or persistence of symptoms [21] that have not been measured in this sample.
Fourth, it has been noted that for within family tests of association, dropped genotypes or genotyping errors for risk alleles with population frequency of less than 0.5 may give rise to false negative findings [49]. Although we were careful to rule out genotyping errors by identifying and re-genotyping ambiguous genotype calls, we did have a high level of dropping genotypes. Both proband and parental genotypes were however in Hardy-Weinberg Equilibrium suggesting that no major genotype biases were affecting this sample. We have attempted to deal with some limitations of within family tests of association by looking for case-control differences for the frequencies of the 2-repeat alleles. Using frequency in Han Chinese controls as 20% from data published by Leung and colleagues [45] and comparing this to the allele frequency of the 2-repeat of our probands (23.8%), there is no significant difference.
One final explanation may be due to hypothesized reciprocity between ADHD associations with the dopamine transporter and DRD4. As hypothesised by Swanson and colleagues [46] the risk allele for ADHD in the dopamine transporter may represent a hyper-efficient variant, whereas the risk alleles for the dopamine D4 receptor may represent a sub-sensitive variant, both leading to a hypo-dopaminergic system. Although additive or epistatic interactions between these two genes are likely, it might also be that they act independently of each other, describing two distinct causes of ADHD. It is therefore possible that samples that have been found to be associated with the DAT1 10-repeat risk allele, may not also be associated to the dopamine receptor D4 risk allele, due to over-representation of one allele or the other in particular populations. For example our UK sample that has also been found to be associated with DAT1 risk variants [47] (Brookes et al, in review), was not found to be associated with the VNTR polymorphism in the dopamine receptor D4 [15]. We investigated this possibility by re-analysing the DRD4 findings separately for individuals homozygous for the dopamine transporter risk allele and those with less than two risk alleles however there was no suggestion of association in these two sub-groups (data not shown).
Conclusions
These results taken together with other reports of Taiwanese and Asian samples find no consistency in the association between genetic variants of DRD4 and ADHD. Further work in Taiwanese and other Asian populations will therefore be required to establish whether the reports of association of DRD4 genetic variants in Caucasian samples can be generalised to Asian populations.
Abbreviations
DRD4 – Dopamine Receptor D4, ADHD – Attention Deficit Hyperactivity Disorder, TDT – Transmission Disequilibrium Test, HHRR – Haplotype-based Haplotype Relative Risk, VNTR – Variable Number Tandem Repeat.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Keeley Brookes and Philip Asherson selected the markers for analysis. Keeley Brookes carried out the genotyping of the population and performed statistical analysis under the supervision of Philip Asherson. DNA was provided by Chih-Ken Chen, Yu-Shu Huang and Yu-Yu Wu. DNA was organised by Xiaohui Xu. Philip Asherson directed the study.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This research was funded by research grants in the UK from the Medical Research Council and the Wellcome Trust and in Taiwan from the Department of Psychiatry, Chan Gung Memorial Hospital and the Taipei City Psychiatric Centre. We thank all families for their participation in this research.
Figures and Tables
Table 1 Allele Frequencies from HHRR analysis and Transmission ratio (T/NT) for TDT. Global and allele-specific tests of association were non-significant.
Case Frequency(%) Control (NT) Frequency(%) T/NT transmission
5' marker
120 bp 0.232 0.236 54 / 55
240 bp 0.768 0.764 179 / 178
Exon 3 VNTR
2 0.234 0.244 46 / 48
3 0.005 0.000 1 / 0
4 0.751 0.736 149 / 146
5 0.010 0.020 2 / 4
Table 2 Haplotype analysis of the two markers did not find either global or haplotype-specific evidence for association.
Haplotype Case (T) Frequency Control (NT) Frequency OR
1–2 17 0.095 12 0.068 1
1–4 21 0.119 26 0.15 0.57
2–2 25 0.144 31 0.176 0.59
2–3 1 0.006 0 0 0.72
2–4 110 0.625 103 0.583 0.77
2–5 2 0.011 4 0.36 0.36
==== Refs
Asherson P Attention Deficit Hyperactivity Disorder in Post-Genomic Era European Journal of Child and Adolescent Psychiatry 2004
LaHoste GJ Swanson JM Wigal SB Glabe C Wigel T King N Kennedy JL Dopamine D4 Receptor Gene Polymorphism Is Association With Attention Deficit Hyperactivity Disorder Molecular Psychiatry 1996 1 121 124 9118321
Swanson JM Sunohara GA Kennedy JL Regino R Fineberg E Wigal T Lerner M Williams L LaHoste GJ Wigal S Association of the Dopamine Receptor D4 (DRD4) Gene With a Refined Phenotype of Attention Deficit Hyperactivity Disorder (ADHD): A Family-Based Approach Molecular Psychiatry 1998 3 38 41 9491811 10.1038/sj.mp.4000354
Smalley SL Bailey JN Palmer CG Cantwell DP McGough JJ Del'Homme MA Asarnow JR Woodward JA Ramsey C Nelson SF Evidence That the Dopamine D4 Receptor Is a Susceptibility Gene in Attention Deficit Hyperactivity Disorder Mol Psychiatry 1998 3 427 430 9774776 10.1038/sj.mp.4000457
Faraone SV Biederman J Weiffenbach B Keith T Chu MP Weaver A Spencer TJ Wilens TE Frazier J Cleves M Sakai J Dopamine D4 Receptor Gene 7-Repeat Allele and Attention Deficit Hyperactivity Disorder Am J Psychiatry 1999 156 768 770 10327912
Sunohara GA Roberts W Malone M Schachar RJ Tannock R Basile VS Wigal T Wigal SB Schuck S Moriarty J Swanson JM Kennedy JL Barr CL Linkage of the Dopamine D4 Receptor Gene and Attention-Deficit/Hyperactivity Disorder J Am Acad Child Adolesc Psychiatry 2000 39 1537 1542 11128331 10.1097/00004583-200012000-00017
Holmes J Payton A Barrett JM Hever T Fitzpatrick H Trumper AL Harrington R McGuffin P Ollier W Worthington J Thapar A A Family-Based Study and Case-Control Study of the Dopamine D4 Receptor Gene and Dopamine Transporter Gene in Attention Deficit Hyperactivity Disorder Molecular Psychiatry 2000 5 523 530 11032386 10.1038/sj.mp.4000751
Holmes J Payton A Barrett J Harrington R McGuffin P Ollier W Worthington J Gill M Kirley A Hawi Z Fitzgerald M Asherson P Curran S Mill J Gould A Taylor E Kent L Craddock N Thapar A Association of DRD4 in Children With ADHD and Comorbid Conduct Problems Am J Med Genet 2002 114 150 153 11857575 10.1002/ajmg.10149
Tahir E Yazgan Y Cirakoglu B Ozbay F Waldman I Asherson PJ Association and Linkage of DRD4 and DRD5 With Attention Deficit Hyperactivity Disorder (ADHD) in a Sample of Turkish Children Mol Psychiatry 2000 5 396 404 10889550 10.1038/sj.mp.4000744
Kotler M Manor I Sever Y Eisenberg J Cohen H Ebstein RP Tyano S Failure to Replicate an Excess of the Long Dopamine D4 Exon III Repeat Polymorphism in ADHD in a Family-Based Study Am J Med Genet 2000 96 278 281 10898899 10.1002/1096-8628(20000612)96:3<278::AID-AJMG8>3.0.CO;2-R
Muglia P Jain U Macciardi F Kennedy JL Adult Attention Deficit Hyperactivity Disorder and the Dopamine D4 Receptor Gene Am J Med Genet 2000 96 273 277 10898898 10.1002/1096-8628(20000612)96:3<273::AID-AJMG7>3.0.CO;2-Z
Eisenberg J Zohar A Mei-Tal G Steinberg A Tartakovsky E Gritsenko I Newmanov L Ebstein RP A haplotype relative risk study of the dopamine D4 receptor (DRD4) exon III repeat polymorphism and Attention Deficit Hyperactivity Disorder (ADHD) J Med Genet 2000 96 258 261 10.1002/1096-8628(20000612)96:3<258::AID-AJMG4>3.0.CO;2-8
Hawi Z McCarron M Kirley A Daly G Fitzgerald M Gill M No Association of the Dopamine DRD4 Receptor (DRD4) Gene Polymorphism With Attention Deficit Hyperactivity Disorder (ADHD) in the Irish Population Am J Med Genet 2000 96 268 272 10898897 10.1002/1096-8628(20000612)96:3<268::AID-AJMG6>3.0.CO;2-#
Payton A Holmes J Barrett JH Hever T Fitzpatrick H Trumper AL Harrington R McGuffin P O'Donovan M Owen M Ollier W Worthington J Thapar A Examining for Association Between Candidate Gene Polymorphisms in the Dopamine Pathway and Attention-Deficit Hyperactivity Disorder: a Family-Based Study Am J Med Genet 2001 105 464 470 11449400 10.1002/ajmg.1407
Mill J Curran S Kent L Richards S Gould A Virdee V Huckett L Sharp J Batten C Fernado S Simanoff E Thompson M Zha J Sham P Taylor E Asherson P Attention Deficit Hyperactivity Disorder (ADHD) and the Dopamine D4 Receptor Gene: Evidence of Association but No Linkage in a UK Sample Molecular Psychiatry 2001 6 440 444 11443530 10.1038/sj.mp.4000881
Manor I Tyano S Eisenberg J Bachner-Melman R Kotler M Ebstein RP The Short DRD4 Repeats Confer Risk to Attention Deficit Hyperactivity Disorder in a Family-Based Design and Impair Performance on a Continuous Performance Test (TOVA) Molecular Psychiatry 2002 7 790 794 12192625 10.1038/sj.mp.4001078
Qian Q Wang Y Li J Yang L Wang B Zhou R Association Studies of Dopamine D4 Receptor Gene and Dopamine Transporter Gene Polymorphisms in Han Chinese Patients With Attention Deficit Hyperactivity Disorder Beijing Da Xue Xue Bao 2003 35 412 418 12947560
Qian Q Wang Y Zhou R Yang L Faraone SV Family-Based and Case-Control Association Studies of DRD4 and DAT1 Polymorphisms in Chinese Attention Deficit Hyperactivity Disorder Patients Suggest Long Repeats Contribute to Genetic Risk for the Disorder Am J Med Genet B Neuropsychiats Genet 2004 128B 84 89 10.1002/ajmg.b.30079
Grady DL Chi HC Ding YC Smith M Wang E Schuck S Flodmen D Spence MA Swanson JM Moyis RK High Prevalence of Rare Dopamine Receptor D4 Alleles in Children Diagnosed With Attention Deficit Hyperactivity Disorder Molecular Psychiatry 2003 8 536 545 12808433 10.1038/sj.mp.4001350
Kustanovich V Ishii J Crawford L Yang M McGough JJ McCracken JT Smalley SL Nelson SF Transmission Disequilibrium Testing of Dopamine-Related Candidate Gene Polymorphisms in ADHD: Confirmation of Association of ADHD With DRD4 and DRD5 Molecular Psychiatry 2004 9 711 717 14699430
El-Faddagh M Laucht M Maras A Vohringer L Schmidt MH Association of Dopamine D4 Receptor (DRD4) Gene With Attention Deficit Hyperactivity Disorder (ADHD) in a High Risk Community Sample: A Longitudinal Study From Birth to 11 Years of Age J Neural Transm 2004 111 883 889 15206004 10.1007/s00702-003-0054-2
Langley K Marshall L van den Bree M Thomas H Owen M O'Donovan M Thapar A Association of the Dopamine D4 Receptor Gene 7-Repeat Allele With Neuropsychological Test Performance of Children With ADHD Am J Psychiatry 2004 161 133 138 14702261 10.1176/appi.ajp.161.1.133
Faraone SV Doyle AE Mick E Biederman J Meta-Analysis of the Association Between the 7-Repeat Allele of the Dopamine D4 Receptor Gene and Attention Deficit Hyperactivity Disorder Am J Psychiatry 2001 158 1052 1057 11431226 10.1176/appi.ajp.158.7.1052
Maher BS Marazita ML Ferrell RE Vanyukov MM Dopamine system genes and Attention Deficit Hyperacitivty Disorder: A meta-analysis Psychiatry Genetics 2002 12 207 215 10.1097/00041444-200212000-00003
McCracken JT Smalley SL McGough JJ Crawford L Del'Homme M Cantor RM Liu A nelson SF Evidence for linkage of a tandem duplication polymorphism upstream of the dopamine D4 receptor gene (DRD4) with Attention Deficit Hyperactivity Disorder (ADHD) Mol Psychiatry 2000 5 531 536 11032387 10.1038/sj.mp.4000770
Arcos-Burgos M Castellanos FX Konecki D Lopera F Pineda D Palacio JD Rapoport JL Berg K Bailey-Wilson J Muenke M Pedigree Disequilibrium Test (PTD) replicates association and linkage between DRD4 and ADHD in multigenerational and extended pedigree from genetic isolate Molecular Psychiatry 2004 9 252 259 15094785 10.1038/sj.mp.4001396
Barr CL Feng Y Wigg KG Schachor R Tannock R Roberts W Malone M Kennedy JL 5'untranslated region of the dopamine D4 receptor gene and Attention Deficit Hyperactivity Disorder Am J Med Genet 2001 105 84 90 11425008 10.1002/1096-8628(20010108)105:1<84::AID-AJMG1068>3.0.CO;2-Q
Todd RD Neuman RJ Cobos EA Jong YS Reich W Heath AC Lack of association of dopamine D4 receptor gene polymorphism with ADHD subtype in a population sample of twins Am J Med Genet 2001 105 432 438 11449395 10.1002/ajmg.1403
Mill JS Caspi A McClay J Sugden K Purcell S Asherson P Craig I McGuffin P Briathwaite A Poulton R Moffitt TE The dopamine D4 receptor and the hyperactivity phenotype: A developmental-epidemiological study Mol Psychiatry 2004 7 383 391 10.1038/sj.mp.4000984
D'Souza UM Russ C Tahir E Mill J McGuffin P Asherson PJ Craig IW Functional Effects of a Tandem Duplication Polymorphism in the 5'Flanking Region of the DRD4 Gene Biol Psychiatry 2004 56 691 697 15522254 10.1016/j.biopsych.2004.08.008
Rogers G Joyce P Mulder R Sellmon D Miller A Allington M Olds r Wells E Kennedy M Association of a duplicated repeat polymorphism in the 5'untranslted region of the DRD4 gene with novelty seeking Am J Med Genet B Neurophsychiatr Genet 2004 126 95 98 10.1002/ajmg.b.20133
Strobel A Lesch KP Jatzke S Paetzold F Brocke B Further evidence for a modulation of novelty seeking by DRD4 Exon III, 5-HTTLPR and COMT Val/Met variants Molecular Psychiatry 2003 8 371 372 12740593 10.1038/sj.mp.4001253
Keltikangas-Jartinen L Elovaninio M Kivimaki M Lichtermann E Ekelund J Peltonen L Association between the type 4 dopamine receptor gene polymorphism and novelty seeking Psychosom Med 2003 65 471 476 12764221 10.1097/01.PSY.0000041547.31072.25
Luciano M Zhu G Kirk KM Whitfield JB Bolter R Heath AC Madden PA Martin NG Effects of Dopamine Receptor D4 variants on alcohol and tobacco use and on novelty seeking: multivariate linkage and association analysis Am J Med Genet 2004 124B 113 123 10.1002/ajmg.b.20077
Wang E Ding YC Flodman P Kidd JR Kidd KK Grady DL Ryder OA Spence MA Swanson JM Moyzis RK The genetics architecture of selection at the human dopamine receptor D4 (DRD4) gene locus Am J Hum Genet 2004 74 931 944 15077199 10.1086/420854
Kluger AN Siegfried Z Ebstein RP A meta-analysis of the association between DRD4 polymorphism and novelty seeking Mol Psychiatry 2002 7 712 717 12192615 10.1038/sj.mp.4001082
Brimer E Levine FM Stimulus seeking behaviour in hyperactive and non hyperactive children J Abnormal Child Psychology 1983 11 131 139 10.1007/BF00912183
Chen CK Chen SL Mill J Huang YS Lin SK Curran S Purcell S Sham P Asherson P The dopamine transporter gene is associated with Attention Deficit Hyperactivity Disorder in a Taiwanese sample Molecular Psychiatry 2003 8 393 396 12740596 10.1038/sj.mp.4001238
Kaufman J Birmaker B Brent D Rau U Flynn C Moneci P Schedule for affective disorders and schizophrenia for school-age children – present and lifetime version (K-SADS-PL): inital reliability and validity data J Am Acad Child and Adolscent Psychiatry 1997 36 980 10.1097/00004583-199707000-00021
Conners CK The Conners rating scale: Instruments for the assessments of childhood psychopathology Duke University 1995
Kahn RS Khoury J Nichols WC Laiphear BP Role of dopamine transporter genotype and maternal prenatal smoking in childhood hyperactivity – impulsive, inattention and oppositional behaviour J Pediatr 2003 143 104 110 12915833 10.1016/S0022-3476(03)00208-7
Maughan B Taylor A Caspi A Moffit TE Prenatal Smoking and Early Childhood Conduct Problems: Testing Genetic and Environment Explanations of the Association Archives of General Psychiatry 2004 61 836 843 15289282 10.1001/archpsyc.61.8.836
West SL Hulsow M Arredondo R Factor Analysis of the Attention Deficit Scales for Adults (ADSA) With a Clinical Sample of Outpatient Substance Abusers Am J Addict 2003 12 159 165 12746091
Ding YC Chi HC Grady DL Morishima A Kidd JR Kidd KK Flodman P Spence MA Schuck S Swanson JM Zhang YP Moyzis RK Evidence of Positive Selection Acting at the Human Dopamine Receptor D4 Gene Locus Proc Natl Acad Sci USA 2002 99 309 314 11756666 10.1073/pnas.012464099
Leung PW Lee CC Hung SF Ho TP Tang CP Kwong SL Leung SY Yuen ST Lieh-Mak F Oosterlaan J Grady D Harxhi A Ding YC Chi HC Flodman P Schuck S Spence MA Moyzis R Swanson J Dopamine Receptor D4 (DRD4) Gene in Han Chinese Children With Attention-Deficit/Hyperactivity Disorder (ADHD): Increased Prevalence of the 2-Repeat Allele Am J Med Genet B Neuropsychiatr Genet 2005 133 54 56 15578612 10.1002/ajmg.b.30129
Swanson JM Flodman P Kennedy J Spence MA Moyzis R Schuck S Murias M Moriarity J Barr C Smith M Posner M Dopamine Genes and ADHD Neurosci Biobehav Rev 2000 24 21 25 10654656 10.1016/S0149-7634(99)00062-7
Curran S Mill J Tahir E Kent L Richards S Gould A Huckett L Sharp J Batten C Fernando S Ozbay F Yazgan Y Simonoff E Thompson M Taylor E Asherson P Association Study of a Dopamine Transporter Polymorphism and Attention Deficit Hyperactivity Disorder in UK and Turkish Samples Mol Psychiatry 2001 6 425 8 11443527 10.1038/sj.mp.4000914
Mann EM Ikeda Y Mueller CW Takahashi A Tao KT Humris E Li BL Chin D Cross-Cultural Differences in Rating Hyperactive-Disruptive Behaviors in Children Am J Psychiatry 1992 149 1539 1542 1415822
Mitchell AA Cutler DJ Chakravarti A Undetected Genotyping Errors Cause Apparent Overtransmission of Common Alleles in the Transmission/Disequilibrium Test Am J Hum Genet 2003 72 598 610 12587097 10.1086/368203
|
16143039
|
PMC1236928
|
CC BY
|
2021-01-04 16:03:33
|
no
|
BMC Med Genet. 2005 Sep 5; 6:31
|
utf-8
|
BMC Med Genet
| 2,005 |
10.1186/1471-2350-6-31
|
oa_comm
|
==== Front
BMC Med GenetBMC Medical Genetics1471-2350BioMed Central London 1471-2350-6-321616229110.1186/1471-2350-6-32Research ArticleAssociation analysis of a highly polymorphic CAG Repeat in the human potassium channel gene KCNN3 and migraine susceptibility Curtain Robert [email protected] James [email protected] Rod [email protected] Mick [email protected] John [email protected] Lyn [email protected] Genomics Research Centre, School of Health Science, Griffith University, Gold Coast, Queensland, Australia2 Institute of Environmental Science and Research, Wellington, New Zealand3 Queensland Clinical Genetics Service, Royal Children's Hospital Health Service District, Brisbane, Queensland, Australia2005 14 9 2005 6 32 32 5 4 2005 14 9 2005 Copyright © 2005 Curtain 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
Migraine is a polygenic multifactorial disease, possessing environmental and genetic causative factors with multiple involved genes. Mutations in various ion channel genes are responsible for a number of neurological disorders. KCNN3 is a neuronal small conductance calcium-activated potassium channel gene that contains two polyglutamine tracts, encoded by polymorphic CAG repeats in the gene. This gene plays a critical role in determining the firing pattern of neurons and acts to regulate intracellular calcium channels.
Methods
The present association study tested whether length variations in the second (more 3') polymorphic CAG repeat in exon 1 of the KCNN3 gene, are involved in susceptibility to migraine with and without aura (MA and MO). In total 423 DNA samples from unrelated individuals, of which 202 consisted of migraine patients and 221 non-migraine controls, were genotyped and analysed using a fluorescence labelled primer set on an ABI310 Genetic Analyzer. Allele frequencies were calculated from observed genotype counts for the KCNN3 polymorphism. Analysis was performed using standard contingency table analysis, incorporating the chi-squared test of independence and CLUMP analysis.
Results
Overall, there was no convincing evidence that KCNN3 CAG lengths differ between Caucasian migraineurs and controls, with no significant difference in the allelic length distribution of CAG repeats between the population groups (P = 0.090). Also the MA and MO subtypes did not differ significantly between control allelic distributions (P > 0.05). The prevalence of the long CAG repeat (>19 repeats) did not reach statistical significance in migraineurs (P = 0.15), nor was there a significant difference between the MA and MO subgroups observed compared to controls (P = 0.46 and P = 0.09, respectively), or between MA vs MO (P = 0.40).
Conclusion
This association study provides no evidence that length variations of the second polyglutamine array in the N-terminus of the KCNN3 channel exert an effect in the pathogenesis of migraine.
==== Body
Background
Migraine is a common, debilitating neurovascular disease characterised by severe recurrent headache, nausea and vomiting, photophobia and phonophobia [1]. It is clinically diagnosed based on criteria specified by the International Headache Society (IHS), defining two major classes of migraine: migraine with aura (MA) and migraine without aura (MO). MA sufferers experience neurovascular disturbances that precede the headache phase of an attack. Although migraine is partly influenced by environmental determinants, there is a significant genetic component, with disease heritability estimated to be up to 60% [2] and mode of transmission multifactorial. The disorder is common with a large Dutch study reporting lifetime prevalence estimates of 33% in women and 13.3% in men [3].
Allelic candidate gene, studies provide the most suitable method for locating genes of small effect contributing to complex genetic disorders, such as migraine [4,5]. In particular, association studies are most powerful when a plausible candidate gene and a sequence variant with potential functional relevance is examined [6].
Mutations in various ion channel genes are responsible for neuromuscular and other neurological disorders. Inherited ion channel mutations or "channelopathies" are increasingly found to be the cause of various neurological disorders in humans (see review [7]). In familial hemiplegic migraine (FHM), a rare subtype of migraine with aura, mutations in the CACNA1A gene (localised at C19p13) have been found (FHM1) [8]. This gene codes for the alpha1A subunit of the neuronal voltage-dependent P/Q-type calcium channel. Recently a second gene, ATP1A2 (FHM2) (localised at C1q23), was implicated in some FHM families [9]. The ATP1A2 ion channel gene, codes for the alpha2 subunit of the Na+, K+ ion ATPase pump. These findings of mutations in these genes have focused attention on central nervous system ionic channels and helped to better understand FHM pathophysiology [10], where the best genetic evidence providing molecular insight into migraine still comes from the mutations detected in the rare form of migraine with aura; FHM [11]. The CACNA1A and ATP1A2 genes have both previously been tested in the common forms of migraine, but no new mutations or the FHM mutations were detected in these MA/MO affected samples [12-14]. Since FHM2-ATP1A2 is partly a potassium channel gene and is localised nearby to the potassium channel KCNN3, it may be interesting to investigate this gene in the common forms of migraine. In general, potassium (K+) channels set the resting membrane potential and regulate the action potential, whereby they control neuronal excitability. The small conductance (SK) calcium (Ca2+) activated K+ channels are responsible for the "after-hyperpolarization" of neurons, which follows a train of action potentials, being activated by the increase in neuronal Ca2+ [15]. The KCNN3 gene (C1q21.3) (a neuronal small conductance calcium-activated potassium channel, localised close to FHM2-ATP1A2;C1q23) encodes a protein of 731 amino acids containing two adjacent polyglutamine arrays (encoded by CAG repeats) in its N-terminal domain separated by 25 amino acids [16]. The first CAG repeat, coding for a polyglutamine stretch in exon 1 at nucleotide 88, also contains a CAA sequence anomaly after 7 repeats of CAG. However this CAG repeat region is not suitable for association studies as it is only slightly polymorphic having usually 12 CAG repeats [15,16]. The second, C-terminal, polyglutamine array, located 111 nucleotides downstream from the initial CAG repeat of the first polyglutamine stretch in exon 1, is highly polymorphic in Caucasian populations with a modal allele length of 19 and a repeat range of 10 to 28 glutamine – CAG repeats [see [15] and [16]]. Small conductance calcium-activated potassium channels (such as KCNN3) play a critical role in determining the firing pattern of neurons via the generation of slow after-polarization and the regulation of intracellular calcium channels [17].
KCNN3, localized at C1q21.3 [15], is positioned close to familial hemiplegic migraine (FHM) type 2, C1q23 [9,18], which is a severe autosomal dominant type of MA [1]. In 1999, Austin et al [19] suggested a mechanistic analogy for the KCNN3 polymorphism may be the small polyglutamine number variations in the calcium channel α1a subunit, encoded by CAG expansions in CACNA1A (a calcium channel implicated in FHM1) [8] which are thought to cause Spinocerebellar ataxia type 6 (SCA6) [20] by loss of channel function mechanism [21].
Polymorphic CAG repeats in the KCNN3 channels affect the regulation of intracellular calcium channels [17]. Since calcium channels regulate numerous processes critical to neuronal function including secretion of neurotransmitters [22], abnormal alterations in calcium channels can cause alterations in the release of neurotransmitters such as serotonin, norepinephrine, and dopamine, which all have been shown to be involved in migraine disease [23-25].
Given that FHM2 maps to C1q23 and KCNN3 localizes nearby at C1q21.3, it may be important to examine the prevalence of the second (highly polymorphic) KCNN3 CAG polymorphism in populations affected and unaffected with migraine. In the current study, we investigated the possibility of an association between migraine (MA and MO affected) and the second (more 3') CAG repeat polymorphism length variation within the KCNN3 gene, using a case-control study of unrelated Australian Caucasian migraine patients and ethnically matched controls.
Methods
Subjects
This study was approved by the Griffith University Ethics Committee and all subjects who participated gave informed consent. The subjects chosen were of Caucasian origin and were categorised based on the diagnostic criteria specified by the International Headache Society (IHS) [1]. Migraine individuals were diagnosed as having migraine with aura (MA) and migraine without aura (MO) from questionnaires and interviews conducted by an experienced clinical neurologist (JM).
In total, 423 DNA samples from unrelated individuals were analysed, of which 202 consisted of migraine patients and 221 non-migraine controls. Of the affected group, 90% of patients had a known family history of migraine, or at least one affected first degree relative. Clinically, the affected group had an average age of approximately 18 yrs disease onset, while average duration of migraine was 20 hours and frequencies of approximately 30 attacks per year. The unaffected control group was recruited from the same geographical location (East Coast of Australia) as the affected group and was matched to the case samples for variables of age (± 5 years), gender and ethnicity (Caucasian). This reduces the possibility of spurious results due to underlying population stratification. Individuals that reported being affected with known migraine comorbid conditions such as mental illness (eg. depression) and cardiovascular disease (eg. stroke) were excluded from the test groups.
Genotyping
Genomic DNA was isolated from whole blood by a standard salting out procedure [26,27]. DNA fragments containing the second KCNN3 CAG polymorphism were amplified by PCR using the oligonucleotides published by Austin et al, 1999 [19]. The sense primer sequence was 5'-CAG CAG CCC CTG GGA CCC TCG C-3', and the anti-sense 5'-GGA GTT GGG CGA GCT GAG ACA G-3'. PCR constituents consisted of the following; 1X buffer J (Master Amp™, EPICENTRE® Technologies), 0.2 μM each of forward and reverse primers, 1 unit of Taq polymerase and 40 ng of genomic DNA, mixed with H20 in a final volume of 10 ul. Thermal cycler parameters consisted of 1 cycle at 95°C for 4 min, for initial denaturation, followed by 35 cycles of denaturation for 1 min at 94°C, primer annealing at 55°C for 1 min and primer extension at 72°C for 30 sec. Final PCR extension consisted of 72°C for 2 min [19].
The DNA samples were genotyped using an ABI PRISM™ 310 Genetic Analyzer and a FAM fluorescent dye labelled forward primer (utilized in the PCR reaction). Raw data was imported into the ABI PRISM™ Genotyper v2.0 DNA Fragment Analysis Software, whereby genotypes were individually called. Independent quality control analysis, performed by a laboratory technician, for the CAG variant consisted of PCR and genotyping of a random selection of 50 cases and controls to test for any differences between initial genotype data.
Statistical analysis
Allele frequencies were calculated from observed genotype counts for the KCNN3 polymorphism. The frequencies for the KCNN3 gene variant were initially assessed for association with migraine using standard contingency table analysis, incorporating the chi-squared test of independence. CLUMP analysis [28], useful for association testing when markers produce sparse contingency tables, was applied to test differences in the allelic distribution between the groups of migraine and controls patients. A two-tailed type I error rate of 5% was chosen for the analyses. Data were analysed by methods described previously, based on the mode of distribution of alleles, with a modal repeat length of 19 CAG repeats [16] and the hypothesis of an association between migraine and larger repeats (>19) analysed by chi-square analysis. Allele sizes were divided into long and short groups with respect to being greater than 19 repeats or equal to and less than 19 CAG repeats in length.
Results
The distribution of allele frequencies for migraine patients compared to controls are displayed in Figure 1, with the distributions of allele frequencies for MA patients, MO patients and controls displayed in Figure 2. Both figures show CAG repeat number along the X-axis versus frequency of distributions (%) displayed along the Y-axis. The repeat range is 12 to 24 repeats in migraine patients, not including the repeat lengths of 22 and 23. For the controls the range is 12 to 26 repeats, except for the 24 CAG repeat size. This displays the highly polymorphic nature of the second CAG expansion within the KCNN3 gene. No significant differences in allelic distributions were observed between the migraine and control groups for standard chi-square analysis (P = 0.090) (Figure 1). Clump analysis, for normal chi-square, also showed no significant difference reported (T1 = 20.111, P = 0.090). However for chi-square clumped 2×2 table analysis (obtained by clumping the columns of the original table to maximise the chi-squared value) [28], between migraine and controls, slightly significant differences were detected within the test data (T4 = 13.001, P = 0.037) (Figure 1). This indicates that there maybe a slight significant difference between the case-control populations with regards to clumped 2×2 table KCNN3 allele distributions. The T3 statistic was also positive for CLUMP analysis (P = 0.025) (Figure 1), although, Sham and Curtis [28] recommends that either normal chi-squared (T1) or chi-squared for "clumped" 2×2 table (T4) analysis should be used. The other two statistics (T2, T3) may lack power to detect association in data sets studied, whereas T1 and T4 perform similarly well [28].
When dividing migraine affected samples into the subtypes MA and MO, distribution of alleles did not differ significantly between MA patients and controls (P > 0.05) or MO patients and controls (P > 0.05) (Figure 2). Analysis was also performed by distributing the alleles into ≤19 (short) and >19 repeats (long) (Table I), described as per Chandy et al, 1998 [16]. The hypothesis of the presence of very long CAG repeats might be expected for a disorder relating to trinucleotide repeat expansion [16]; this was tested by chi-square analysis. Migraine probands showed no significant difference in allelic length distribution of the CAG repeats compared with controls (P = 0.15). Though the number of long CAG repeats (>19 repeats) was higher in the migraine group (92 compared to 83 for controls, Table 1) and the incidence of the CAG short repeat (<19 repeats) was higher in the control (359) group than in the migraine (312) group. Similarly, the allele frequency distributions for the migraine subgroups, MA and MO compared to controls did not approach significance (P = 0.46 and P = 0.09, respectively), or between MA vs MO (P = 0.40) (Table I).
Discussion
Polyglutamine disorders are due to CAG repeat expansions that cause a toxic gain of function of mutant expanded proteins, in which protein misfolding, interference with DNA transcription and RNA processing, activation of apoptosis and dysfunction of cytoplasmic elements have all been invoked in the toxic process. [29]. These CAG expansions have been involved in a number of neurological diseases, including Huntington's disease (HD), Dentatorubralpallidoluysian atrophy (DRPLA), Kennedy's disease and Spinocerebellar ataxia (SCA) 1–3, 6–7, 12 and 17 [29].
Migraine is a polygenic multifactorial disease influenced by genetic and lifestyle characteristics. At present the mode(s) of inheritance is unclear and the type and number of migraine genes involved in the disease is not known. Since both FHM and familial typical migraine (FTM) display some clinical overlap [1], it is thought that the more prevalent FTM may also be a channelopathy [30]. The CACNA1A gene, involved with FHM1 [8] controls a number of fundamental neuronal processes including mediation and release of neurotransmitters such as serotonin [31]. The KCNN3 gene acts to regulate the firing pattern of neurons via the generation of slow after-polarization and the regulation of intracellular calcium channels [17]. Patients with migraine have autonomic nervous system dysfunction and altered levels of neurotransmitters. Therefore, decreases in activity and regulation of intracellular calcium channels, that may be caused by CAG repeat expansions in the KCNN3 gene, would result in decreased levels of neurotransmitters (such as serotonin) and autonomic nervous system dysfunction. These states appear to be analogous to those found during migraine headache in that decreases in plasma levels of serotonin would essentially act to constrict the arterioles, therefore causing dilation of the larger arteries, possibly causing pain [32,33].
Wittekindt et al, 1998 [15] reported that the KCNN3 gene is a good candidate for schizophrenia and bipolar disorder (BD), as well as for other neurological disorders. This includes migraine, a painful neurological disease that affects 33% of women and 13.3% of men [3].
However, even though there was a slight indication of significance in one analysis (clump T4 statistic) [28], overall the results of this study proved that there was no statistically significant association between allelic frequencies of migraine and non-migraine patients. Also no significant difference in allelic frequencies was observed in the migraine subtypes, MA and MO when compared to the control population. There have been many contradicting studies that have shown a significant over-representation or non-significant representation of long CAG repeats in the KCNN3 gene in patients with schizophrenia and bipolar disease compared to population controls [see [34]]. In a meta-analysis of association studies for schizophrenia and bipolar disorder with the CAG-repeat length in KCNN3, the results demonstrated that the risks for both of these disorders were largely, if not entirely, independent of the CAG-repeat in the KCNN3 gene [34]. This study examined CAG repeat lengths (the second, C-terminal, polyglutamine array in exon 1 of KCNN3) in migraine patients and also found no significant evidence to suggest long CAG repeats (>19) are over-represented in patients with migraine compared to controls, even though there was a higher incidence reported for long CAG repeat (>19 repeats) in the migraine group compared to the controls (Table I), it did not reach statistical significance (P < 0.05). However, a recent study by Mossner et al, 2005 [35], reported an association between the second highly polymorphic polyglutamine stretch and migraine. In this study the authors found allele 15 to be associated with migraine (P = 0.027; total migraine) [35]. Allele 15 is extremely rare with a 1.6% frequency in MA, 1.5% in MO and a 0.2% control frequency found in Mossner et al's study [35]. In our investigation of the second KCNN3 – CAG repeat we found a frequency in allele 15 of 0.9% in MA (only 2 out of 222 alleles), 1.65% in MO (3 out of 182 alleles) and 0.9% in controls (4 out of 442 alleles) (see distribution of alleles, Figures 1 and 2). From the two studies it can be seen that this allele is too rare to make any assumptions of whether it is migraine related. A population base far greater in number is needed in order to obtain meaningful data from allele 15. Also, the fact that this study utilised samples that were all carefully diagnosed following IHS guidelines [1], including all samples from the migraine unaffected-control population and Mossner et al, 2005 [35] did not, leaves their positive results questionable. Mossner et al had made a migraine-unaffected assumption with more than half their controls (119/232 samples). These individuals that were utilised as controls, were an undiagnosed, anonymous blood donor group [35], that were not directly interviewed or diagnosed according to the International Headache Society guidelines [1]. This could have a tremendous impact on the outcomes of their study, due to the rarity of allele 15 in the second CAG repeat of exon1 in KCNN3.
The present association study was conducted in carefully diagnosed (according to IHS guidelines [1]), age, sex and ethnicity matched case-control populations. However the results provided no evidence that KCNN3 gene confers susceptibility to the subtypes of migraine (MA and MO), or common migraine.
Conclusion
We conclude that our data does not confer with Mossner et al, 2005 [35], providing no evidence that a variation at the polymorphic, second CAG repeat locus, in the KCNN3 gene, influences susceptibility to migraine, or to the migraine subtypes, MA and MO.
List of abbreviations
KCNN3 neuronal small conductance calcium-activated potassium channel, subfamily N, member 3, MA migraine with aura, MO migraine without aura.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RC carried out the optimisation of this molecular genetic study and drafted the manuscript. JS carried out PCR and genotyping. RL carried out statistical analysis. MO carried out DNA sample preparation for the populations. JM carried out patient clinical diagnosis. LG 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.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Figures and Tables
Figure 1 Distribution of allele frequencies by CAG repeat number for the second CAG polymorphism of the KCNN3 gene observed in migraine and control groups.
Figure 2 Distribution of allele frequencies by CAG repeat number for the second CAG polymorphism of the KCNN3 gene observed in migraine subtypes (MA and MO) and control groups.
Table 1 Long (>19 repeats) and short (≤19 repeats) alleles: CAG repeat number in KCNN3; analysis performed utilising the comparison of long and short allele repeat numbers between migraine and control populations.
Allele Size
Group Short (≤19rpts) Long (>19rpts)
MA 175 47
MO 137 45
Migraine (Total) 312 92
Control 359 83
* Analysis of long and short alleles: no significant differences were obtained between migraine vs control groups (P = 0.15), MA vs controls (P = 0.46), MO vs controls (P = 0.09), or between MA vs MO (P = 0.40)
==== Refs
Headache Classification Subcommittee of the International Headache Society The international classification of headache disorders: 2nd edition Cephalalgia 2004 24 9 160 14979299
Ulrich V Gervil M Kyvik KO Olesen J Russell MB The inheritance of migraine with aura estimated by means of structural equation modelling J Med Genet 1999 36 225 7 10204850
Launer LJ Terwindt GM Ferrari MD The prevalence and characteristics of migraine in a population-based cohort: the GEM study Neurology 1999 53 537 542 10449117
Owen MJ McGuffin P Association and linkage: complimentary strategies for complex disorders J Med Genet 1993 30 638 639 8411048
Risch N Merikangas K The future of genetic studies of complex human diseases Science 1996 273 1516 1517 8801636
Nöthen MM Propping P Fimmers R Association versus linkage studies in psychosis genetics J Med Genet 1993 30 634 637 8411047
Kullmann DM Hanna MG Neurological disorders caused by inherited ion-channel mutations Lancet Neurol 2002 1 157 66 12849484 10.1016/S1474-4422(02)00071-6
Ophoff RA Terwindt GM Vergouwe MN van Eijk R Oefner PJ Hoffman SM Lamerdin JE Mohrenweiser HW Bulman DE Ferrari M Haan J Lindhout D van Ommen GJ Hofker MH Ferrari MD Frants RR Familial hemiplegic migraine and episodic ataxia type2 are caused by mutations in the Ca2 + channel gene CACNL1A4 Cell 1996 87 543 52 8898206 10.1016/S0092-8674(00)81373-2
De Fusco M Marconi R Silvestri L Atorino L Rampoldi L Morgante L Ballabio A Aridon P Casari G Haploinsufficiency of ATP1A2 encoding the Na+/K+ pump alpha-2 subunit associated with familial hemiplegic migraine type 2 Nature Genet 2003 33 192 196 12539047 10.1038/ng1081
Fumal A Schoenen J [Genetics of migraines: from ionic channels to single nucleotide polymorphisms?] Rev Med Liege 2004 59 367 77 15346975
Wessman M Kaunisto MA Kallela M Palotie A The molecular genetics of migraine Ann Med 2004 36 462 73 15513297 10.1080/07853890410018060
Lea RA Curtain RP Hutchins C Brimage PJ Griffiths LR Investigation of the CACNA1A gene as a candidate for typical migraine susceptibility Am J Med Genet 2001 105 707 12 11803518 10.1002/ajmg.1609
Jen JC Kim GW Dudding KA Baloh RW No mutations in CACNA1A and ATP1A2 in probands with common types of migraine Arch Neurol 2004 61 926 8 15210532 10.1001/archneur.61.6.926
Curtain RP Lea RA Tajouri L Haupt L Ovcaric M MacMillan J Griffiths L Analysis of Chromosome 1 Microsatellite Markers and the FHM-2 (ATP1A2 Gene) Mutations in Migraine With and Without Aura Pedigrees Neurological Research 2005 accepted 27/02/05 16157018
Wittekindt O Jauch A Burgert E Schärer L Holtgreve-Grez H Yvert G Imbert G Zimmer J Hoehe M Macher J Chiaroni P Calker D Crocq M Morris-Rosendahl D The human small conductance calcium-regulated potassium channel gene (hSKCs3) contains two CAG repeats in exon 1, is on chromosome 1q21.3, and shows a possible association with schizophrenia Neurogenetics 1998 1 259 265 10732800 10.1007/s100480050038
Chandy KG Fantino E Wittekind O Kalman K Tong L Ho T-H Gutman GA Crocq M-A Fargus JJ Isolation of a novel potassium channel gene hSKCa3 containing a polymorphic CAG repeat: a candidate gene for schizophrenia and bipolar disorder Mol Psychiatry 1998 3 32 37 9491810 10.1038/sj.mp.4000353
Köhler M Hirschberg B Bond CT Kinzie JM Marrion NV Adelman JP Small-conductance, calcium-activated potassium channels from mammalian brain Science 1996 273 1709 1714 8781233
Keryanov S Gardner KL Physical mapping and characterization of the human Na, K-ATPase isoform, ATP1A4 Gene 2002 292 151 66 12119109 10.1016/S0378-1119(02)00647-9
Austin CP Holder DJ Ma L Mixson LA Caskey CT Mapping of KCNN3 to chromosome 1q21 and investigation of linkage of CAG repeat polymorphism to schizophrenia Mol Psychiatry 1999 4 261 266 10395216 10.1038/sj.mp.4000548
Zhuchenko O Bailey J Bonnen P Ashizawa T Stockton DW Amos C Dobyns WB Subramony SH Zoghbi HY Lee CC Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the alpha 1A-voltage-dependent calcium channel Nat Genet 1997 15 62 9 8988170 10.1038/ng0197-62
Frontali M Spinocerebellar ataxia type 6: channelopathy or glutamine repeat disorder? Brain Res Bull 2001 56 227 31 11719255 10.1016/S0361-9230(01)00574-3
Flink MT Atchison WD Ca2+ channels as targets of neurological disease: Lambert-Eaton Syndrome and other Ca2+ channelopathies J Bioenerg Biomembr 2003 35 697 718 15000529 10.1023/B:JOBB.0000008033.02320.10
Villalon CM Centurion D Valdivia LF de Vries P Saxena PR Migraine: pathophysiology, pharmacology, treatment and future trends Curr Vasc Pharmacol 2003 1 71 84 15320857 10.2174/1570161033386826
Peroutka SJ Migraine: a chronic sympathetic nervous system disorder Headache 2004 44 53 64 14979884 10.1111/j.1526-4610.2004.04011.x
Lea RA Dohy A Jordan K Quinlan S Brimage PJ Griffiths LR Evidence for allelic association of the dopamine beta-hydroxylase gene (DBH) with susceptibility to typical migraine Neurogenetics 2000 3 35 40 11085595
Blin N Stafford DW Isolation of high molecular-weight DNA Nucleic Acids Res 1976 3 2303 987581
Miller SA Dykes DD Plensky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acid Res 1988 16 1215 3344216
Sham P Curtis D Monte Carlo tests for associations between disease and alleles at highly polymorphic loci Ann Hum Genet 1995 59 97 105 7762987
Everett CM Wood NW Trinucleotide repeats and neurodegenerative disease Brain 2004 127 2385 405 15329351 10.1093/brain/awh278
Ophoff RA van den Maagdenberg AM Roon KI Ferrari MD Frants RR The impact of pharmacogenetics for migraine Eur J Pharmacol 2001 413 1 10 11173058 10.1016/S0014-2999(00)00949-3
Codignola A Tarroni P Clementi F Pollo A Lovallo M Carbone E Sher E Calcium channel subtypes controlling serotonin release from human cell lung carcinoma cell line J Biol Chem 1993 268 26240 26247 8253745
Ferrari MD Migraine Lancet 1998 351 1043 51 9546526 10.1016/S0140-6736(97)11370-8
Goadsby PJ Zagami AS Lambert GA Neural processing of cardiovascular pain: a synthesis of the central structures involved in migraine Headache 1991 31 365 371 1889975 10.1111/j.1526-4610.1991.hed3106365.x
Glatt SJ Faraone SV Tsuang MT CAG-repeat length in exon 1 of KCNN3 does not influence risk for schizophrenia or bipolar disorder: a meta-analysis of association studies Am J Med Genet B Neuropsychiatr Genet 2003 121 14 20 12898569 10.1002/ajmg.b.20048
Mossner R Weichselbaum A Marziniak M Freitag CM Lesch KP Sommer C Meyer J A highly polymorphic poly-glutamine stretch in the potassium channel KCNN3 in migraine Headache 2005 45 132 6 15705118 10.1111/j.1526-4610.2005.05027.x
|
16162291
|
PMC1236929
|
CC BY
|
2021-01-04 16:03:33
|
no
|
BMC Med Genet. 2005 Sep 14; 6:32
|
utf-8
|
BMC Med Genet
| 2,005 |
10.1186/1471-2350-6-32
|
oa_comm
|
==== Front
BMC Med EducBMC Medical Education1472-6920BioMed Central London 1472-6920-5-321616228710.1186/1472-6920-5-32Research ArticleExposure to the field of renal transplantation during undergraduate medical education in the UK Edwards Anusha G [email protected] Alex [email protected] Justin D [email protected] Department of Surgery, Southmead Hospital, Westbury on Trym, Bristol BS10 5NB, UK2005 14 9 2005 5 32 32 6 7 2005 14 9 2005 Copyright © 2005 Edwards et al; licensee BioMed Central Ltd.2005Edwards 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
There is a lack of surgeons in the field of renal transplantation, with a predicted shortage of over 20 consultants by the year 2005. Early positive exposure to the field, commencing at undergraduate level, has been identified as being vital to improving rates of recruitment. This study was performed to assess the exposure of undergraduates to the field of renal transplantation during medical education in the UK.
Methods
In October 2004 a questionnaire was sent to the clinical deans of all UK medical schools regarding undergraduate exposure to renal transplantation.
Results
Twenty-five replies were received, giving a response rate of 96%. All but one school had a centre for renal transplantation in their region. Three schools (12%) gave no formal lecture or tutorial on the subject during the entire course. Of the remainder, between one to four formal sessions were provided, ranging from 15 minutes to 3 hours duration.
Six medical schools (24%) provided no compulsory clinical exposure to renal transplantation, with a further five (20%) saying that students may receive exposure by chance. The average length of attachment was three weeks. Twenty-one medical schools (84%) provided between 1–10% of students a choice to study renal transplantation, as part of electives and special study modules.
Conclusion
This study reveals a variation between, and within, medical schools in the levels of formal teaching. If the trends in recruitment to renal transplantation are to be reversed, we have an obligation to improve upon the medical education that students currently receive.
==== Body
Background
Renal transplantation is the first choice treatment for many patients with end-stage renal failure, with the potential for improved quality of life and increased life expectancy [1,2]. However, in the UK, the speciality suffers from a lack of qualified surgeons with a predicted shortage, in 1999, of over twenty consultant renal transplant surgeons by the year 2005 [3]. The most recent figures have shown that of the 94 renal transplant consultant posts in the UK, 12 are filled by locums [4].
Previous work has highlighted the multiple factors that deter surgical trainees from the speciality, which include a lack of exposure to the speciality, at an early stage during training [3]. This has lead to a call for the inclusion of transplantation within basic surgical training rotation programmes.
Last year a crisis meeting, organised by the British Transplantation Society, (BTS) was held to address the problem with recruitment into renal transplantation. It was identified that early, positive exposure to the field should be commenced at undergraduate level and a number of ways of addressing this were suggested [5]. A recent study, at a single UK medical school demonstrated a lack of exposure to and knowledge of renal transplantation [6], but no literature exists regarding the national situation. Therefore, this survey was conducted to assess how much exposure to the field of transplantation medical students in the UK are currently receiving.
Methods
In October 2004 a questionnaire was sent to the clinical deans of all 26 UK medical schools. It consisted of eight questions regarding exposure of undergraduates to the field of renal transplantation. (Figure 1) Although not validated, it was based upon a questionnaire from a similar study in the field of ENT [7]. Non-responders were sent a further copy of the questionnaire after four weeks, which were followed with a telephone call after a further four weeks, if necessary.
Figure 1 Questionnaire distributed to clinical deans of all UK medical schools.
Results
Replies were received from twenty-five medical schools, giving a response rate of 96%. However, one response was incomplete as the clinical dean felt that the course was "too integrated" to be analysed in detail.
Three schools (12%) gave no formal lecture or tutorial on the subject during the entire undergraduate course. Of the remainder, between one to four formal sessions were provided, ranging from 15 minutes to 3 hours duration (Figure 2). Two of these medical schools specified that formal sessions were only available to students attached to a transplantation firm, but gave no indication of what proportion of students this applied to.
Figure 2 A plot of the number of formal teaching sessions provided by UK medical schools.
Despite the fact that all but one of the responding medical schools had a centre for renal transplantation in their region six medical schools (24%) provided no compulsory clinical exposure to renal transplantation, with a further five (20%) saying that students may receive exposure by chance. The average length of attachment was three weeks, ranging from one to five weeks. However, twenty-one medical schools (84%) provided between 1–10% of students a choice to study renal transplantation, usually as part of electives and special study modules. The duration of such options ranged from two to four weeks.
The majority of medical schools, fifteen, (60%), included the teaching of renal transplantation with other specialities, whilst one school felt that the course was too integrated to tell. (table 1)
Table 1 The number of medical schools teaching renal transplantation in conjunction with a range of other subjects
Subject Number of medical schools
Renal medicine 8
Surgery 3
Urology 2
Pathology 1
Immunology 4
Transplant medicine 1
Ethics 2
Regarding the assessment of knowledge of renal transplantation, one medical school always (4%), twelve schools sometimes (48%) and nine never (36%) formally assesses it during examinations.
Fifteen of the responding clinical deans (60%) felt that there was adequate exposure to renal transplantation in the undergraduate curriculum. Of the remainder, one thought there wasn't, two were unsure, three found it difficult to find teachers and three thought that it is, and should remain a postgraduate subject.
Discussion
This study reveals a worrying variation both between, and within, medical schools in the levels of formal teaching and exposure to renal transplantation. There is no data available regarding the correlation between level of exposure to transplantation and final career choice. However, previous work has demonstrated that one factor that deters surgical trainees from a career in transplantation is a lack of exposure to the speciality [3]. We therefore believe that if the trends in recruitment to the field are to be reversed, everyone working in the speciality has an obligation to improve upon the medical education that students currently receive in the UK.
This work is the first formal assessment of exposure to renal transplantation in UK medical schools. The excellent response rate makes this a comprehensive study. However, the fact that one response was incomplete suggests that with integrated courses assessing the exposure to one speciality can be difficult. Although the questionnaire was not formally validated, it was designed to be used in today's medical schools, and was based upon a questionnaire used in a similar study assessing undergraduate exposure to ENT surgery [8].
A previous study performed at a single medical school, with two local centres of renal transplantation, demonstrated variability in the amount of exposure to transplantation and worryingly low levels of knowledge about the field, amongst final year medical students [6]. This national study suggests that similar or lower levels of exposure are occurring throughout the country, which has worrying implications for the future of renal transplantation, both in terms of recruitment and organ procurement.
It has been shown that increased knowledge about organ donation is associated with an increased likelihood of holding an organ donor card and feeling more comfortable in approaching relatives of potential organ donors [9]. In the UK the speciality also suffers from an ever-increasing discrepancy between the number of organs donated and the number of patients on the transplant waiting list [10]. Over the last few years non heart-beating programmes have been introduced in the UK, in order to increase the number of organs available. They can potentially increase the transplant rate by 20–40% [11]. However, if the doctors of tomorrow are not aware of such programmes and do not feel equipped to approach relatives, levels of donation are unlikely to be maximised.
Obviously, there are other factors that deter trainees from a career in transplantation [3] and these too must be tackled in order to reverse the current recruitment trends. These include the significant out-of-hours commitment and appropriate recompense, both of which are currently being tackled within the frameworks of the European Working Time Directive and the new consultant contracts respectively. Career progression and training are also issues of concern for trainees, which have been partly dealt with by the recent provision of funding by the Department of Health for more specialist registrar training posts in renal transplantation [4].
One clinical dean suggested that the F2 year might be an opportunity to give students a "taster" in the speciality. This proposal has also been considered by the BTS, with an F2 year, consisting of 4 months each of renal transplantation, nephrology and general surgery or urology. The BTS is currently liasing with the national Modernising Medical Careers group about this option and individual transplant units have been encouraged to submit appropriate bids to their Postgraduate Dean.
The recommendations made by the BTS crisis meeting for early positive exposure at the undergraduate level included a presence at medical school careers fairs and surgical societies, along with the introduction of a full time educator at UK Transplant [12]. Such measures should help to raise the profile of the speciality. Indeed studies in the USA have demonstrated that positive encounters with surgeons can favourably influence the perceptions of first year medical students towards a career in surgery [13]. However, an early interest must be fostered throughout undergraduate training if students are to seriously consider transplantation as a career. For many doctors this comes from actually witnessing the speciality at work, receiving good, practical teaching [14] or being inspired by an individual in the field [15]. Therefore, the authors believe that as teaching time is limited during undergraduate studies medical schools should ensure that the person chosen to teach renal transplantation is someone who is able to communicate well with students and can present the speciality in such a way as to inspire. This individual does not necessarily have to be a transplant surgeon as the subject can also be taught in conjunction with other subjects such as renal medicine and immunology. In addition to conventional bedside teaching and lectures the multi disciplinary nature of transplantation should be exploited. Involving students in patient education days and the writing of information sheets or patient web pages may prove more memorable and inspirational to some students than conventional forms of exposure.
Conclusion
This study has demonstrated that amongst medical schools in the UK there is considerable variation in the levels of exposure to the field of renal transplantation. Given the current shortage of renal transplant surgeons this is of great concern. In order to reverse the recruitment crisis all professionals involved in the speciality should take it upon themselves to increase exposure to and profile of the field.
List of abbreviations
UK – United Kingdom
BTS – British Transplantation Society
ENT – Ear, nose and throat
USA – United States of America
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AE was involved in study design and execution and manuscript preparation.
AN was involved in questionnaire distribution and collection and manuscript preparation.
JM was involved in study design and manuscript preparation.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Laupacis A Keown P Pus N Krueger H Ferguson B Wong C Muirhead N A study of the quality of life and cost-utility of renal transplantation Kidney Int 1996 50 235 242 8807593
Wolfe RA Ashby VB Milford EL Ojo AO Ettenger RE Agodoa LY Held PJ Port FK Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant N Engl J Med 1999 341 1725 1730 10580071 10.1056/NEJM199912023412303
McGrath JS Shehata M Attitudes of surgical trainees towards transplantation surgery as a career Transpl Int 1999 12 303 306 10551994 10.1007/s001470050232
Rigg K Hard graft rewarded BMA News 2005 Saturday 16th April 14
British Transplantation Society. The Crisis in Recruitment to Kidney Transplant Surgery 2004
Edwards AG Weale AR Morgan JD A survey of medical students to assess their exposure to and knowledge of renal transplantation BMC Medical Education 2004 32 15617572 10.1186/1472-6920-4-32
Mace AD Narula AA Survey of current undergraduate otolaryngology training in the United Kingdom J Laryngol Otol 2004 118 217 220 15068520 10.1258/002221504322928008
Schaeffner ES Windisch W Freidel K Breitenfeldt K Winkelmayer WC Knowledge and attitude regarding organ donation among medical students and physicians Transplantation 2004 77 1714 1718 15201671 10.1097/00007890-200406150-00015
Transplant Activity Report. 2000
Cho YW Terasaki PI Cecka JM Gjertson DW Transplantation of kidneys from donors whose hearts have stopped beating N Engl J Med 1998 338 221 225 9435327 10.1056/NEJM199801223380403
Kozar RA Lucci A Miller CC Azizzadeh A Cocanour CS Potts JR Fischer CP Brundage SI Brief intervention by surgeons can influence students toward a career in surgery J Surg Res 2003 111 166 169 12842462 10.1016/S0022-4804(03)00104-5
Ko CY Escarce JJ Baker L Klein D Guarino C Predictors for medical students entering a general surgery residency: National survey results Surgery 2004 136 567 572 15349103 10.1016/j.surg.2004.05.021
Thakur A Fedorka P Ko C Buchmiller-Crair TL Atkinson JB Fonkalsrud EW Impact of mentor guidance in surgical career selection J Pediatr Surg 2001 36 1802 1804 11733910 10.1053/jpsu.2001.28842
|
16162287
|
PMC1236930
|
CC BY
|
2021-01-04 16:30:56
|
no
|
BMC Med Educ. 2005 Sep 14; 5:32
|
utf-8
|
BMC Med Educ
| 2,005 |
10.1186/1472-6920-5-32
|
oa_comm
|
==== Front
BMC MicrobiolBMC Microbiology1471-2180BioMed Central London 1471-2180-5-501613732210.1186/1471-2180-5-50Methodology ArticleExpression capable library for studies of Neisseria gonorrhoeae, version 1.0 Brettin Thomas [email protected] Michael R [email protected] Ying [email protected] Roxie M [email protected] Alexandra [email protected] Laura [email protected] Chris [email protected] Thomas J [email protected] Jason [email protected] Heather [email protected] Nathan J [email protected] Susan [email protected] Kimberly [email protected] Susi [email protected] Erica [email protected] Joseph P [email protected] George [email protected] Magdalene [email protected] Cindy Grove [email protected] Bioscience Division, Los Alamos National Laboratory, Los Alamos, NM 87545, USA2 Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI 48824-4320, USA3 Department of Molecular Microbiology and Immunology, Oregon Health and Science University, Portland, OR 97201-3098, USA4 Human Genome Sequencing Center, Baylor College of Medicine, Houston TX 77030, USA5 Department of Medical Microbiology and Immunology, University of Wisconsin Medical School, Madison, WI 53706, USA6 Leicester Warwick Medical School, University of Warwick, Coventry, UK2005 1 9 2005 5 50 50 28 1 2005 1 9 2005 Copyright © 2005 Brettin et al; licensee BioMed Central Ltd.2005Brettin 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 sexually transmitted disease, gonorrhea, is a serious health problem in developed as well as in developing countries, for which treatment continues to be a challenge. The recent completion of the genome sequence of the causative agent, Neisseria gonorrhoeae, opens up an entirely new set of approaches for studying this organism and the diseases it causes. Here, we describe the initial phases of the construction of an expression-capable clone set representing the protein-coding ORFs of the gonococcal genome using a recombination-based cloning system.
Results
The clone set thus far includes 1672 of the 2250 predicted ORFs of the N. gonorrhoeae genome, of which 1393 (83%) are sequence-validated. Included in this set are 48 of the 61 ORFs of the gonococcal genetic island of strain MS11, not present in the sequenced genome of strain FA1090. L-arabinose-inducible glutathione-S-transferase (GST)-fusions were constructed from random clones and each was shown to express a fusion protein of the predicted size following induction, demonstrating the use of the recombination cloning system. PCR amplicons of each ORF used in the cloning reactions were spotted onto glass slides to produce DNA microarrays representing 2035 genes of the gonococcal genome. Pilot experiments indicate that these arrays are suitable for the analysis of global gene expression in gonococci.
Conclusion
This archived set of Gateway® entry clones will facilitate high-throughput genomic and proteomic studies of gonococcal genes using a variety of expression and analysis systems. In addition, the DNA arrays produced will allow us to generate gene expression profiles of gonococci grown in a wide variety of conditions. Together, the resources produced in this work will facilitate experiments to dissect the molecular mechanisms of gonococcal pathogenesis on a global scale, and ultimately lead to the determination of the functions of unknown genes in the genome.
==== Body
Background
Neisseria gonorrhoeae (gonococcus), a Gram-negative diplococcus, is one of two pathogenic members of the Neisseriaceae family of bacteria. N. gonorrhoeae is the causative agent of the sexually transmitted disease, gonorrhea, one of the oldest documented infectious diseases. Gonorrheal disease has significant morbidity both in the US and worldwide. According to the Centers for Disease Control [1], >350,000 cases of gonorrhea were reported in the United States in 2002. The World Health Organization [2] estimates that over 19 million cases occur annually in the African continent alone. Treatment of gonorrhea is increasingly problematic due to the high frequency of acquisition of resistance to multiple antibiotics [3,4] and to the observation that gonococcal infection does not elicit protective immunity [5]. Gonorrheal infections, though not usually life-threatening, also enhance the transmission of HIV [6].
N. gonorrhoeae is strictly a human pathogen, with no known animal reservoir. The bacterium has no environmental niche, and cannot survive outside the human host. In adults, N. gonorrhoeae is acquired primarily through sexual contact. However, the eyes of newborn infants may be infected by passing through an infected birth canal, resulting in the condition, ophthalmia neonatorum, which can lead to blindness. In most cases, gonococcal infections are limited to the urogenital tract, causing urethritis in men and cervicitis in women. Occasionally, gonococci cross the epithelial barrier to enter the bloodstream causing septicemia, and transit to the joints resulting in arthritis. In women, ascending infections from the endocervix can result in pelvic inflammatory disease, salpingitis, tubal blockage and infertility. N. gonorrhoeae can also establish a carrier state in which apparently healthy individuals harbor culturable and infectious bacteria [7]. Carriers are thought be important for disease dissemination. A recent study revealed that the gonococcal carriage rate in women was 6.7% in a major metropolitan area [8].
Due to the importance of N. gonorrhoeae to human health, much research effort has focussed on identifying virulence factors and elucidating the biochemical interactions of these factors with the host cell [9-11], with the goal of developing vaccines and alternative treatments. It is clear, however, that in order to fully understand the capabilities of this organism to cause disease and elude eradication, it will be necessary to ultimately determine the functions of a great deal more of the gene products encoded by the gonococcal genome. The recent genome sequencing makes possible a variety of genomic and proteomic studies of N. gonorrhoeae. To facilitate such studies, we have cloned into a bacteriophage lambda-based recombination cloning system (Gateway® [12], Invitrogen, Carlsbad, CA) 1624 of the 2189 predicted ORFs from the genome of N. gonorrhoeae strain FA1090 [13], and 48 of the 61 ORFs of the gonococcal genetic island (GGI) of strain MS11 [14,15]. This clone-set allows the generation of transcriptional and translational fusions without the necessity of additional cloning and sequencing. Coupled to the construction of this clone set, DNA microarrays were generated by spotting the insert DNA onto glass slides. Preliminary experiments with the clone set and DNA arrays indicate that this system is suitable for studies of expression of genes from N. gonorrhoeae in heterologous systems as well as for the study of global gene expression in this organism.
Results
Design of oligonucleotide primers
The goal of this project was to create a plasmid library representing the annotated ORFs of N. gonorrhoeae. The Gateway® Cloning System from Invitrogen [12] was selected for several reasons. First, Gateway® uses a recombination-based cloning method which has the added benefit that once an archival clone is sequence-validated, subsequent recombinants (ie. into expression vectors) do not need to be sequenced. Second, the initial clones lack transcriptional machinery such that the cloned ORFs are not expressed, thus avoiding problems from lethality due to troublesome gene products. Third, there are several expression and epitope-tagging vector options for the subsequent study of proteins encoded by the cloned ORFs, allowing a variety of approaches to studying their functions. The high efficiency also lends itself to high throughput approaches that are suitable for automation.
A total of 2071 unique primer pairs were successfully designed for the 2189 annotated ORFs of the FA1090 genome [13] and 61 ORFs of the GGI [14,15]. These primers were gene specific, and their termini contain sequences for recombination cloning into the entry vector, pDONR221 (Invitrogen). All primers were designed such that the final recombination product yielded the native start codon at the 5' end of the gene, including the 206 of the 229 predicted ORFs with alternative (non-ATG) start codons. Since nearly half of the genes of this group (110/229) are annotated (that is encode putative proteins with significant similarity to proteins of known function), it is very possible that they are functional genes in the gonococcus and were thus included in the clone set design. Whether or not they encode functional proteins will ultimately depend on the results of future expression and mutation studies. The 23 ORFs of this subset not included were less than 400 bp in length and considered too small for the Gateway® system (see below). For the 149 ORFs encoding predicted proteins with an amino-terminal signal sequence (as identified by PSORT [16] and SignalP [17]), sequences encoding the signal sequence were removed and an ATG start codon placed at the 5' end of the remaining coding sequence. This was done to reduce problems of expression of hydrophobic signal sequences and to facilitate future expression and targeting studies for such recombinant proteins.
The primer design strategy was iterative, starting with an annealing temperature range of 62°–72°C and primer length set to 18 nt. All ORFs were included in the first iteration, and those ORFs for which a primer pair was not selected were subjected to subsequent iterations. In each subsequent iteration the primer length parameter was increased by one, up to a maximum primer length of 34 nt. Primers larger than this were not designed in part due to cost and convenience of oligo synthesis in our 96-well format. Next, the annealing temperature range was expanded 5°C in both the positive and negative directions and each primer size from 18 to 34 nt was tried again. No further iterations were attempted after the annealing temperature range exceeded 47°–87°C. Failing to meet either of these criteria resulted in a primer pair not being designed for the ORF. There were a total of 177 ORFs for which no primers were designed (see Additional file 2), 174 from FA1090 and 3 from the GGI. Most of these ORFs (165/177 = 93%) were less than 500 bp in length, and were not included since the Gateway® system is reportedly less efficient for cloning fragments of this small size.
To each of the gene specific primer sequences for the 5' ends of the ORFs was added a 21 nt sequence including a consensus ribosome binding site (Shine-Dalgarno sequence). To each of the gene specific primer sequences for the 3' ends of the ORFs was added 20 nt corresponding to the 3' end of the attB2 site necessary for recombination into pDONR221. These were the primary PCR primers, the sequences of which are available on request. A single pair of primers was then designed for a secondary amplification to generate gene specific products containing the attB sites for the recombinatory cloning step. The 5' secondary primer contained the 24 nt attB1 tail and the 21 nt sequence (Shine-Dalgarno) common to all of the primary 5' primer sequences. The 3' secondary primer contained the remaining 10 nt of the attB2 site and the 20 nt attB2 sequence common to all of the primary 3' primer sequences. The attB sequences were as recommended in the Gateway® manual.
PCR amplification of ORFs
The first round of PCR, using gene specific primers, included a total of 2071 different primer pairs. Plates 1 and 2 were organized as pilot reactions and contained primers designed to amplify products ranging from 143 bp to 3485 bp in length. Genomic DNA from N. gonorrhoeae strain FA1090 was used as a template for the reactions. Agarose gel analysis of the amplicons showed that 173 of the 192 reactions yielded a product of the expected size, an efficiency of 90%. Plates 3–21 were then arranged with increasing size of expected product, with plate 3 containing the smallest products and plate 21 the largest. Several primer pairs (from plates 1 and 2) were included in plates 3–21 as internal controls. Plate 22 was an additional control plate, with one half of the plate (rows A-D) duplicated on the other half (rows E-H). Plate 23 corresponded to the genes of the GGI present in strain MS11 [14,15], and MS11 genomic DNA was used as the template for this plate of reactions. 5 μl of each reaction from the first round of PCR of each plate were run on agarose gels and scored for production of a product and whether it was of the expected size. The results showed 89% of the reactions to produce a product of the correct size.
Following the primary amplification with gene specific primers, all products (in the original 96-well format) were diluted 1:100 and an aliquot subjected to a secondary amplification using a pair of primers corresponding to the sequences common to all of the amplicons, and including additional sequences to generate a complete attB site for the cloning reaction. Aliquots of the secondary amplification were then used directly in the recombination reaction to generate entry clones.
Construction of the library
Secondary amplicons were inserted into pDONR221 by in vitro recombination between the attB sites introduced at the ends of the amplicons and the attP sequences of the vector, maintaining the 96-well format arrangement. Cloning reactions were then transformed into E. coli strain DH5α and a portion of the transformation mix plated on LB plates containing kanamycin. Individual transformants were screened by PCR to determine the presence of and size of the insert. For the first round of screening, four independent transformants from each reaction were screened, maintaining the original 96-well format. A product of 350 bp was observed if no insert was present, a positive clone was identified as having a product 350 bp larger than the size of the corresponding primary PCR product.
For the initial set of 2147 transformation reactions, clones corresponding to 1165 genes were identified in the first four transformants screened, an efficiency of 54%. This efficiency varied greatly with the predicted size of insert, the smallest inserts (plate 3) were 83% positive in the first four screened and the next largest inserts (plate 20) had 17% positive in the first four screened. Plate 21, which had the largest inserts, only yielded 3 transformants as positive after several rounds of screening. Additional transformants for those clones not identified in the first round of screening were individually cultured, screened by PCR, and positive clones frozen down as they were identified. This approach yielded an additional 354 clones.
Following the initial rounds of screening, a list of missing clones was generated and the amplification and cloning steps repeated, optimizing several parameters and analyzing on an individual basis. This approach yielded an additional 283 clones, for a total of 1802 which were subsequently sequenced. Arrangement of the clones in plates for the master set and for sequencing was on an "as identified" basis, such that they are not arranged as in the original 96-well format. Each of the 21 plates contain viable clones in up to 95 of the 96 available positions, with position H12 (and additional wells on some plates) left empty for controls, providing a unique identity for several of the plates.
Sequence verification of the clones
Transformants identified as having an insert of the predicted size were grown in 96-well plates and DNA isolated for sequence analysis. DNA isolation and sequencing was done by at the Human Genome Sequencing Center at Baylor College of Medicine (HGSC) using the same set of primers used to screen transformants for insert size. Sequence reads were posted onto an HGSC website and subsequently downloaded by FTP. DNA reads were processed initially using the STADEN DNA analysis software package [18]. Binary files were converted into .exp files using PREGAP4 to generate text files for each individual sequence read, with sequence quality cutoffs. These data are provided in Additional File 1 (AF1 exp sequence files.zip), and a description of the labelling scheme for the files is in the Methods section.
The data were next analyzed by BLAST [19] against the N. gonorrhoeae genome sequence database [13]. Clones expected to contain inserts from the GGI, not present in FA1090, were analyzed by BLAST 2 [20] using sequence of the individual GGI ORFs ([15], GenBank accession number AY803022). BLAST results were then manually tabulated in a file containing the expected gene for each archived clone. Of the 1802 sequenced clones, 58 were expected to be duplicates, leaving 1744 unique clones expected in the clone set. 1550 of the sequences were readable and corresponded to a predicted ORF from N. gonorrhoeae, 1399 of them unique, corresponding to 151 duplicates. Some duplicates were expected, and the remainder likely due to cross contamination from neighboring wells. Of those sequence validated, 55 were not in positions predicted. Most of these were due to human error, such as obvious well transpositions and numbering transpositions. 26 of these, however, had inserts in a backwards orientation and were incomplete. These clones will not be usable in subsequent recombination reactions using the Gateway® system. Together, these data indicate that at present we have a collection of 1672 individual clones from N. gonorrhoeae, 48 of which are from the GGI, and 83% of which have been sequence validated. A list of genes in the clone set with sequencing result information can be found in Additional File 2 (AF2 NG clone set seq status.xls).
Overexpression of randomly chosen ORFs
In order to examine the flexibility of using the clone set to construct various derivatives for which the Gateway® system was designed, three randomly selected pDONR221 derivatives were used to create inducible glutathione-S-transferase (GST) fusions. Plasmids containing ORFs NG1490 (aspS, encodes aspartyl-tRNA synthetase), NG1561 (xthA, encodes exodeoxyribonuclease III), and NG1641 (pivNG, encodes a pilin gene inverting homolog, PivNG) are predicted to encode native proteins of 9.6, 29.0, and 36.2 kdal respectively. Plasmid DNA of pDONR221 derivatives containing these three genes were recombined with the destination vector pDEST15, an N-terminal GST fusion vector. The resulting recombinants were then transformed into E. coli BL21-AI for expression analysis. SDS-PAGE analysis of the proteins after a 2 hr induction with 0.2% L-arabinose is shown in Figure 1. The results show high levels of induction for each of the fusion proteins, with the sizes as predicted (GST adds 29 kdal).
Figure 1 SDS-PAGE analysis of NG ORF-GST fusions. Equivalent amounts of total protein was boiled in sample buffer and electrophoresed on 10% polyacrylamide gels. Gels were stained with Coomassie Blue. Lane 1: MW markers; lane 2: NG1490(AspS)-GST, uninduced; lane 3: NG1490 induced (38.6 kdal); lane 4: NG1561(XthA)-GST uninduced; lane 5: NG1561 induced (58.0 kdal); lane 6: NG1641(PivNG)-GST uninduced; lane 7: NG1641 induced (65.2 kdal).
DNA microarray production
Since a very small portion of the secondary amplicons were used for the cloning reactions (5 μl of a 50 μl reaction), the remaining products were used to produce a set of DNA arrays for gene expression analysis. Gel electrophoresis (data not shown) indicated that the efficiency after the secondary amplification was 81%, representing 1681 ORFs of a total of 2071. To generate a more complete DNA array, a second set of primers (360) were designed to amplify internal portions of those ORFs not visible following the secondary PCR. Primers to amplify sequences of two small RNAs were also designed: NG0892.1, ffs, encodes the 4.5S RNA component of the gonococcal signal recognition particle [21]; and NG0880.1, tmRNA, encodes an RNA that tags abnormal proteins in the cell arising from stalled ribosomes and targets them for proteolysis [22]. Gel analysis of the amplicons produced from the internal PCR primers showed 294 of the 362 to produce products of the expected size, for an efficiency of 82%. Together, the cloning amplicons (1681) and internal ORF and RNA amplicons (294) represent a minimum of 1975 ORFs of the N. gonorrhoeae genome. This is a minimal estimate since all products (regardles of gel result) were to be spotted, and some products might be present, but at amounts too low to be visualized. DNA samples were processed and spotted onto glass slides as described in Methods. As a first test, the arrays were hybridized with a Cy3-labelled random nonamer oligonucleotide (Qiagen). A scan of the slide at 532 nm showed spots at the appropriate positions where DNA had been spotted, and blank spots at the buffer control spots. The next test was to hybridize the arrays with labelled genomic DNA. Total DNA from N. gonorrhoeae strain MS11 was digested with RsaI and labelled by including Cy3-dCTP in a random-primed Klenow DNA polymerase reaction (Roche Applied Science, Indianapolis, IN). DNA arrays were hybridized and then scanned at 532 nm. Valid hybridization signals (at least one standard deviation above background) were detected for 98% of the spots expected to contain DNA, with those of the MS11-specific island comparable in intensity to those of FA1090 amplified DNA. Overall, the results showed valid signals for 2035 individual genes, with several in duplicate. The fact that this number is higher than expected based on agarose gel analysis of the amplicons before spotting (see above) indicates that some of the reactions produced a product, but at amounts too low to be visualized by ethidium bromide (EtBr) staining. Thus, these arrays represent 90% (2035/2250) of the predicted ORFs of N. gonorrhoeae (including 58 ORFs of the GGI; [14,15]) and 98% of those for which primers were designed.
Discussion
Despite the advent of antibiotics in the 1940's, disease due to infection with N. gonorrhoeae remains a major health problem worldwide. The reasons for this are multi-fold. First, resistance to antibiotics by N. gonorrhoeae continues to rise [1,3,4]. In addition, treatment with high levels of broad spectrum antibiotics (which is frequently done since patients often do not return for follow up treatment) kills many bacteria of the (often beneficial) normal flora as well as the disease-causing microbe. Second, there is an incredibly high frequency of asymptomatic gonococcal infection, occurring in 5–10% of infected men and up to 50% of infected women. This represents a major reservoir for transmission of the infection. Furthermore, undiagnosed and untreated gonococcal salpingitis can lead to fallopian tube blockage. Partial blockage can result in ectopic pregnancy, which can be life threatening, and complete blockage of the fallopian tubes often leads to infertility. Third, development of a vaccine to protect against gonorrhea has been seriously hampered by the observation that gonococcal infection does not elicit protective immunity [5]. Patients can be reinfected following treatment, and can even be infected by multiple strains at a given time. Thus, alternative treatments and preventative strategies for gonococcal infection are sorely needed.
As a first step in the identification of such alternative treatments and preventatives, it will be necessary to more thoroughly understand the biology of the gonococcus and the molecular mechanisms involved in its interactions with the host environment. Much of the studies to date have focussed primarily on identifying the molecules on the surface of the bacterium that directly interact with the host, and the toxic moieties involved in damage to host cells. Many of the molecules identified are outer membrane (OM) components [23-27], and iron utilization proteins (reviewed in [28]). A few are secreted [29], or shed in blebs [30-32]. There has also been significant work in the identification of eukaryotic host cell receptors for gonococcal surface proteins (reviewed in [10]).
Interactions between gonococci and epithelial cells are beginning to be unraveled. Chen and Clark showed that contact with Hec-1-B human endocervical epithelial cells increases gonococcal infectiveness and that the process involves, in part, de novo protein synthesis by the bacterium [33]. The gonococcal type IV pilus (Tfp) and Opa proteins promote attachment, invasion and trans-epithelial trafficking. The mechanisms underlying Tfp- and Opa-mediated virulence are not yet understood, but these surface structures modulate a series of events in the infected epithelial cell, among them Ca2+ fluxes [34,35], cortical rearrangements [36], and receptor phosphorylation [37-39]. Tfp retraction enhances the activation of stress-responsive kinases and the transcription of cytoprotective genes in the infected cell [40], and triggers the infected cell to produce a molecule that alters bacterial motility behavior [41]. Finally, binding of gonococci to primary urethral cells up-regulates anti-apoptotic factors [42]. These and other observations indicate that gonococcal infection requires the active participation of both the bacterium and the host cell.
The recent completion of the annotated genome sequence of N. gonorrhoeae [13,43], coupled with the development of high throughput methods for the analysis of gene expression and function, provide an opportunity to significantly advance the study of gonococcal biology and pathogenesis. Like many sequenced genomes, nearly half (44%) of the genes of the annotated gonococcal genome encode hypothetical proteins of unknown function. Furthermore, many of the annotations are based on homologies at the nucleotide and/or amino acid level, and the actual function of the gonococcal proteins have not been demonstrated. In order to realize the full potential of information gleaned from the genome sequence of this (and any) organism, it will be necessary to assign functions to all of the genes of the genome. The newly emerging fields of functional genomics and proteomics offer much promise towards achieving the goal of eventual assignment of functions for each and every gene in a given organism.
In this work, we describe the initial phases of the construction of an expression-capable clone set representing the annotated ORFs of the gonococcal genome using a recombination-based cloning system. The advantages of the system used for this set are numerous. 1) The original sequences in the clone set contain only the ORFs, not the gene expression sequences, thus avoiding the issue of expression-related lethality of the recombinants. 2) The clones can be transferred to a number of expression systems (prokaryotic and eukaryotic), allowing the regulation of genes for overproduction of proteins, or the production of proteins out of the context of the particular environments so as to study their functions. 3) The clones can also be transferred to vectors that result in epitope fusions, such as hexa-histidine, GST, green fluorescent protein (GFP), Lumio™, etc., to the proteins of interest. Protein fusions are useful in localizing proteins (within the bacterium or infected cell), in determining protein-interacting partners, in allowing smaller step purification protocols for structural and activity studies, and for antibody production. This ability has been demonstrated by constructing IPTG-inducible GST-fusions from three random clones from this set (Fig. 1). 4) The clone set is also catalogued in such a way that individual clones of interest are easily identified and recovered from the clone bank (see Additional file 2). 5) Entry clones can also be used to create knockouts by in vitro transposition [44,45] or shuttle mutagenesis [46] followed by transformation into naturally competent gonococci [47]. This system is also ammenable to automation, thus increasing the potential output and consistency in the data obtained.
The N. gonorrhoeae clone set thus far includes 1672 of the 2250 predicted ORFs of the genome [13], of which 83% are sequence-validated. Included in this set are 48 of the 61 ORFs of the MS11 GGI [14,15]. While this clone set is not yet complete, we believe these initial efforts have resulted in generating a valuable resource for the Neisseria research community. It is hoped that others in the community will share compatible reagents and add to the clone set, making it more comprehensive over time.
Coupled to the clone set construction, a PCR-amplicon based DNA microarray was generated. DNA microarrays are a powerful tool that allow one to measure relative transcript levels for essentially each gene of the genome simultaneously. These DNA arrays represent 2035 ORFs of the N. gonorrhoeae genome: 1977 from strain FA1090 [48] and 58 from the MS11 GGI [14], comprising 90% of the genes of the genome. Preliminary studies show that these arrays are suitable for examining global gene expression in N. gonorrhoeae.
Many bacterial pathogens are known to respond to changes in their physical environment, often integrating responses to several environmental signals via complex regulatory networks to control expression of a variety of genes [49-53]. The examples of regulatory systems characterized in gonococci are few, with the best characterized being the response to iron availability [54,55] and antimicrobial compounds [56,57]. The advent of microarray technologies has opened avenues of research on global gene expression in both prokaryotes and eukaryotes, providing opportunities for studying a variety of organisms, including such genetically intractable microbes as Trepanema pallidum [58] and Chlamydia trachomatis [59]. Thus, the use of DNA arrays will allow us to more fully explore the response of N. gonorrhoeae to environmental signals at the gene expression level. Since genes are typically only transcribed when the gene product function is required, expression profiles and cluster analyses will allow us to begin to determine the functions of unknown genes in the genome. Thus far, use of these DNA arrays has led to the identification of a regulator involved in the modulation of gonococcal gene expression upon adherence to epithelial cells in [45].
In summary, the tools described in this work represent a resource which will facilitate experiments to dissect the molecular mechanisms of gonococcal pathogenesis on a global scale. Combining these tools with the gonococcal infection models (tissue culture [60], organ culture [61-63], and the mouse model [64]), will allow us to make significant advances in the study of this important pathogen, thus providing us with the knowledge necessary to design therapeutics with which to treat and prevent gonococcal disease.
Methods
Gene predictions and initial sequence preparation
All ORF IDs for strain FA1090 reference records can be found at the STDGEN Neisseria gonorrhoeae annotated genome sequence database [13]. All gene sequences were prepared to include the natural start and stop codons. NG0540 and NG0634 have internal stop codons in the sequence database, and nothing was done to correct for this. Sequences of the ORFs of the MS11 GGI [15] have been deposited in the GenBank database with the accession number AY803022.
Signal peptide identification
In order to determine whether the ORFs contained signal sequences, the programs PSORT [16] and SignalP [17] were employed. If both programs predicted a cleavable signal peptide for a given ORF sequence, that constituted "high support". The result of the signal peptide analyses showed 149 sequences with high support for a signal peptide. For those gene nucleotide sequences with a high support signal peptide, the nucleotide sequence representing the signal peptide was removed prior to cloning primer design. When both programs predicted different cleavage sites, the cleavage site that represented the shorter signal peptide was chosen. The rationale for this choice was that it would be better to include a bit of the signal peptide in the PCR product than to exclude a bit of the mature protein in the PCR product.
Primer design
In the first cycle, the forward and reverse primers were fixed at the same length. This was due to the ease at which Primer3 [65] could be used. The input file was all gene sequences for which no signal peptide sequence was detected (see above). For the second cycle, the prim.aux file was manually inspected. This file contained information about successful primer picks where only a left or right primer could be picked. The strategy was to combine primers of different length. The potential for primer-dimer formation using this strategy was also assessed.
N.g. ORF cloning
Primers for each of the genes were purchased from Illumina, Inc (San Diego, CA) and were designed to add sequences corresponding to part of the attB site necessary for recombination into the Gateway® entry vector, pDONR221 (Invitrogen). Primary amplification was done using genomic DNA at a concentration of ~10 ng/reaction and gene specific primers at 2.5 μM. Reaction mix contained dNTPs, reaction buffer, MgCl2 and Taq polymerase as recommended by the manufacturer (Roche). Reactions conditions for primary PCR were as follows: denaturation at 94°C for 10 min; 10 cycles of 94°C 30 sec, 50°C 1 min, 74°C 1–5 min (depending on length of predicted product); 20 cycles of 94°C 30 sec, 55°C 1 min, 74°C 1–5 min (depending on length of predicted product); and a final extension at 74°C for 10 min. Following the primary amplification with the primer set, products were diluted 1:100 and 1 μl used as template for a secondary amplification using a pair of primers corresponding to the partial attB site common to all of the amplicons, and including additional sequences to generate a complete attB site for the cloning reaction. Reactions conditions for secondary PCR were as follows: denaturation at 94°C for 1 min; 5 cycles of 94°C 15 sec, 45°C 30 sec, 68°C 2 min; 15 cycles of 94°C 15 sec, 55°C 30 sec, 68°C 2 min. 5 μl of the 50 μl PCR was removed and used for cloning, and the remainder used for agarose gel analysis and printing of DNA arrays (N.g.array version 1.0, see below). Cloning reactions were performed according to the manufacturer's instructions (Invitrogen), transformed into E. coli strain DH5α, and transformants selected for kanamycin resistance. Individual transformants were picked in to wells of 96-well plates containing 100 μl L broth containing kanamycin (50 mg/l) and the same toothpick then used to place a small amount of bacteria directly into another plate containing a PCR cocktail. PCR was done using the M13 universal primers, which flank the att sites of the entry vector, pDONR221. A product of 350 bp was observed if no insert was present, providing an internal control for the PCR reactions. Individual clones were identified, stocked in duplicate, and grown for DNA isolation for sequencing.
Sequencing
Sequencing runs from each end of the insert of each of the clones was determined to verify the ORF inserted. Complete sequence verification, (ie. both strands completely across the insert) was not done as it was determined to be impractical. DNA sequencing reactions were performed at the Baylor College of Medicine HGSC. Additional File 1 is a zipped file containing each of the sequence reads as .exp files generated using PREGAP4, and can be opened and read using word processing software. The file is separated into folders labelled SeqPlate #, which refers to sequencing plate number (1–21), and corresponds to the SP# designation in the list of clones in Additional File 2. SP18 sequencing reactions were done twice (SeqPlate 18-1, SeqPlate 18-2) as a sequencing reaction control, and the reactions of SP5 were analyzed three times (SeqPlate 5-1, SeqPlate 5-2, SeqPlate 5-3) as controls. SP20 does not exist, SP22 was not sent for sequencing, and the quality of SP17 sequence was too poor to be readable. The naming of the individual read files is as follows: BGACA(project code) # (1, 2, or 3 = reaction) D or F (primer D = forward, F = reverse) # (box, not same as SP) # (01–96; well, 1 = A1, 2 = B1, 9 = A2, and so on) A or B (run). For example: BGACA3D1701A (found in SeqPlate 1 file) corresponds to SP1 well A1 sequenced with the forward primer (5' end of ORF), result from the third reaction and the first gel.
Gene expression analysis
pDONR221 derivatives chosen from the clone set were recombined with the destination vector, pDEST15 (Invitrogen), an N-terminal GST fusion vector. Plasmid DNA from the pDONR221 derivatives was isolated and incubated with pDEST15 in a LR recombination reaction performed according to the manufacturer's instructions. The recombination mixes were transformed into E. coli DH5α and transformants were selected on LB plates containing 50 mg/l Carbenicillin (Cb50). To test the possibility of false positive clones, the overexpression clones were tested for growth on chloramphenicol (20 mg/l), on which expression recombinants should not grow since the chloramphenicol resistance gene of pDEST15 is replaced by the insert. The resulting plasmids were purified and transformed into E. coli BL21-AI (Invitrogen) which expresses T7 RNA polymerase from the araBAD promotor, and transformants were selected on LB Cb50 plates. BL21-AI strains containing the pDEST15 derivatives were grown overnight and used to inoculate fresh LB medium containing Cb50 to an OD600 of 0.05. Expression of the GST-fusion proteins in E. coli BL21-AI was induced at an OD600 of 0.5 by addition of L-arabinose to a final concentration of 0.2%. Aliquots were removed after 2 hr and total proteins electropheresed on 10% polyacrylamide SDS gels [66].
DNA microarray construction
PCR amplicons remaining from the cloning reactions (described above) were concentrated and then spotted in duplicate onto TeleChem SuperAmine glass slides (TeleChem International, Inc., Sunnyvale, CA) using a GeneMachines Omnigrid 100 (GeneMachines, Inc., San Carlos, CA) with 16 TeleChem Chipmaker 3 pins at the Genome Technology Support Facility (GTSF) at Michigan State University. Preparation of probes and hybridization conditions were as described [45]. Hybridized microarray slides were scanned using a GenePix 4000B scanner (Axon Instruments, Union City, CA) and images were processed and analyzed using GenePix version 4.1 software.
Availability of clone set
The master library was used to generate a limited set of stock plates, which are stored at various locations as reference stocks. An additional set is stored at Michigan State University and is regarded as the working copy, which is the sole source of material for distribution. Interested parties should contact Cindy Arvidson (corresponding author) to arrange for a copy of the clone set. The clones will only be available as an intact set, individual clones or groups of clones are not available.
Availability of microarrays
DNA arrays will be generated by re-amplification of the amplicons used for the recombination-cloning reactions using primers recognizing the common sequences at the ends of the amplicons at the MSU GTSF. Interested parties should contact Cindy Arvidson (corresponding author) to request slides, which will be produced upon request.
Authors' contributions
Thomas Brettin: design of clone set construction, primer design
Michael R. Altherr: design of clone set construction, design and supervision of high-throughput PCR amplification of ORFs and recombination cloning into entry vectors
Ying Du: construction and tests of DNA microarrays, amplification and cloning of several ORFs not obtained in the initial batch cloning
Roxie M. Mason: cataloging of all cloning reactions processed at MSU (Arvidson lab), screening of transformants and preparation of clones for sequencing
Alexandra Friedrich: recombination of selected clones into expression vectors and analysis of proteins produced (Figure 1)
Laura Potter: initial project design, screening and cataloging of all cloning reactions processed at OHSU (So lab), screening of transformants and preparation of clones for sequencing
Chris Langford: batch BLAST analysis of sequencing results
Thomas J. Keller: batch BLAST analysis of sequencing results, wrote new program to faciliate analysis of sequence data
Jason Jens: processing of sequence data, batch program design for analysis of subset of clones from Gonococcal Genetic Island
Heather Howie: screening of transformants and preparation of clones for sequencing
Nathan J. Weyand: screening of transformants and preparation of clones for sequencing
Susan Clary: screening of transformants and preparation of clones for sequencing, amplification and cloning of several ORFs not obtained in the initial batch cloning
Kimberly Prichard: primer design
Susi Wachocki: high-throughput PCR amplification of ORFs and cloning into entry vectors
Erica Sodergren: sequencing of clone set
Joseph P. Dillard: contributed data of the Gonococcal Genetic Island prior to publication, participated in project design
George Weinstock: initial project design
Magdalene So: initial project design, supervison of all work performed at OHSU, substantial writing of manuscript
Cindy Grove Arvidson: initial project design, supervison of all work performed at MSU, screening of transformants from LANL, cloning of several ORFs not cloned in initial high-throughput cloning, collated and organanized all sequence data, primary writing of manuscript, corresponding author
Supplementary Material
Additional File 2
Microsoft Excel file. Sheet 1: SP compiled – lists all clones sequenced and the results of the sequence analysis with respect to the gene expected and the gene identified by BLAST analysis of the sequence read. Sheet 2: entire ORF set – complete list of ORFs as annotated at [13] the status of each with respect to cloning primers designed, clone identified, and sequence verification.
Click here for file
Additional File 1
Zipped file which includes all of the sequencing data as .exp files generated from binary data using PREGAP4. These .exp files contain all of the sequences with quality cutoffs included and can be opened with word processing programs.
Click here for file
Acknowledgements
Work in the Arvidson laboratory was supported by start-up funds to C.G.A. from the Colleges of Human and Osteopathic Medicine at Michigan State University, and an Intramural Research Grant from the MSU Vice President for Research and Graduate Studies. Work performed at the Los Alamos National Laboratory was supported by the Molecular Foundations of Pathogenesis project, U.S. Department of Energy Contract W-7405-ENG-36 and by funds from the LANL Laboratory Directed Research Development Program to M.R.A. Work in the So laboratory was supported in part by NIH grants AI34560, AI47260 and AI49973 to M. S. Alexandra Friedrich was supported by a fellowship within the Postdoc-Programme of the German Academic Exchange Service (DAAD).
The authors are extremely grateful to thank Jeffrey Landgraf of the Michigan State University GTSF for help and advice in the production and use of DNA arrays and Robert Britton of the MSU Department of Microbiology and Molecular Genetics for advice and the use of his array scanner and software. We also thank Audrey Butcher and Catherine Beauduy (Arvidson laboratory) for their participation in analyzing potential clones for the clone set and Anita Amin, Judy Hernandez, Donna Muzny, and the Baylor College of Medicine Human Genome Sequencing Center for assitance in sequencing the clone set.
==== Refs
Centers for Disease Control and Prevention
World Health Organization
Fox KK Knapp JS Holmes KK Hook EW Judson FN Thompson SE Washington JA Whittington WL Antimicrobial resistance in Neisseria gonorrhoeae in the United States, 1988-1994: the emergence of
decreased susceptibility to the fluoroquinolones. J Infect Dis 1997 175 1396 1403 9180179
Shlaes D Levy S Archer G Antimicrobial resistance: new directions. ASM News 1991 57 455 458
Schneider H Hale TL Zollinger WD Seid RCJ Hammack CA Griffiss JM Does an experimental gonococcal infection protect human volunteers from subsequent reinfection? Abst Tenth Intnl Path Neisseria Conf 1996 10 13
Laga M Manoka A Kivuvu M Malele B Tuliza M Nzila N Goeman J Behets F Batter V Alary M Heyward WL Ryder RW Piot P Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS 1993 7 95 102 8442924
Fletcher JD Stratton J Chandler CS Sparling PF Epidemiologic notes and reports penicillin-resistant gonorrhea -- North Carolina. MMWR 1983 32 273 275 6405170
Turner CF Rogers SM Miller HG Miller WC Gribble JN Chromy JR Leone PA Cooley PC Quinn TC Zenilman JM Untreated gonococcal and chlamydial infection in a probability sample of adults. JAMA 2002 287 726 733 11851539 10.1001/jama.287.6.726
Hauck CR Meyer TF 'Small' talk: Opa proteins as mediators of Neisseria-host-cell communication. Curr Opin Microbiol 2003 6 43 49 12615218 10.1016/S1369-5274(03)00004-3
Merz AJ So M Interactions of pathogenic Neisseriae with epithelial cell membranes Annu Rev Cell Dev Biol 2000 16 423 457 11031243 10.1146/annurev.cellbio.16.1.423
Pujol C Eugene E Morand P Nassif X Do pathogenic neisseriae need several ways to modify the host cell cytoskeleton? Microbes Infect 2000 2 821 827 10955963 10.1016/S1286-4579(00)90367-8
Walhout AJ Temple GF Brasch MA Hartley JL Lorson MA van den Heuvel S Vidal M GATEWAY recombinational cloning: application to the cloning of large numbers of open reading frames or ORFeomes. Meth Enzymol 2000 328 575 592 11075367
Brettin T Neisseria gonorrhoeae annotated genome sequence database.
Dillard JP Seifert HS A variable genetic island specific for Neisseria gonorrhoeae is involved in providing DNA for natural transformation and is found more often in disseminated infection isolates. Mol Microbiol 2001 41 263 277 11454218 10.1046/j.1365-2958.2001.02520.x
Hamilton HL Dominguez NM Schwartz KJ Hackett KT Dillard JP Neisseria gonorrhoeae secretes chromosomal DNA via a novel type IV secretion system. Mol Microbiol 2005 55 1704 1721 15752195 10.1111/j.1365-2958.2005.04521.x
Nakai K Kanehisa M Expert system for predicting localization sites in Gram-negative bacteria. Proteins 1991 11 95 110 1946347 10.1002/prot.340110203
Nielsen H Engelbrecht J Brunak S von Heijne G Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering 1997 10 1 6 9051728 10.1093/protein/10.1.1
STADEN DNA sequence analysis package
Altschul SF Madden TL Schäffer AA Zhang J Zhang Z Miller W Lipman DJ Gapped BLAST and PSI-BLAST: a new generation of protein database search
programs. Nucl Acids Res 1997 25 3389 3402 9254694 10.1093/nar/25.17.3389
Tatusova TA Madden TL Blast 2 sequences - a new tool for comparing protein and nucleotide sequences FEMS Microbiol Lett 1999 174 247 250 10339815 10.1016/S0378-1097(99)00149-4
Frasz C Arvidson CG A role for both DNA and RNA in GTP hydrolysis by the Neisseria gonorrhoeae signal recognition particle receptor. J Bacteriol 2003 185 801 808 12533455 10.1128/JB.185.3.801-808.2003
Gueneau de Novoa P Williams KP The tmRNA website: reductive evolution of tmRNA in plastids and other symbionts. Nucleic Acids Res 2004 32 D104 108 14681369 10.1093/nar/gkh102
Fischer SH Rest RF Gonococci possessing only certain P.II outer membrane proteins interact with human neutrophils. Infect Immun 1988 56 1574 1579 3131247
Punsalang AP Sawyer WD Role of pili in the virulence of Neisseria gonorrhoeae. Infect Immun 1973 8 255 263 4200266
Rudel T Scheurerpflug I Meyer TF Neisseria PilC protein identified as type-4 pilus tip-located adhesin Nature 1995 373 357 359 7830772 10.1038/373357a0
Swanson J Studies on gonococcus infection. IV. Pili: their role in attachment of gonococci to tissue culture cells J Exp Med 1973 137 571 589 4631989 10.1084/jem.137.3.571
van Putten JP Duensing TD Carlson J Gonococcal invasion of epithelial cells driven by P.IA, a bacterial ion channel with GTP binding properties J Exp Med 1998 188 941 952 9730895 10.1084/jem.188.5.941
Schryvers AB Stojiljkovic I Iron acquisition systems in the pathogenic Neisseria Mol Microbiol 1999 32 1117 1123 10383753 10.1046/j.1365-2958.1999.01411.x
Simpson DA Hausinger RP Mulks M Purification, characterization, and comparison of the immunoglobulin A1 proteases of Neisseria gonorrhoeae. J Bacteriol 1988 170 1866 1873 3127382
Gregg CR Melly MA Hellerqvist CG Coniglio JG McGee ZA Toxic activity of purified lipopolysaccharide of Neisseria gonorrhoeae for human fallopian tube mucosa. J Infect Dis 1981 143 432 439 6785365
Griffiss J Schneider H Mandrell R Yamasaki R Jarvis G Kim J Gibson B Hamadeh R Apicella M Lipooligosaccharides: The principal glycolipids of the Neisserial outer membrane. Rev Infect Dis 1988 10 S287 S295 2460911
Melly MA McGee ZA Rosenthal RS Ability of monomeric peptidoglycon fragments from Neisseria gonorrhoeae to damage human fallopian-tube mucosa. J Infect Dis 1984 149 378 386 6425421
Chen JCR Bavoil P Clark VL Enhancement of the invasive ability of Neisseria gonorrhoeae by contact with Hec1B, an adenocarcinoma endometrial cell line. Mol Microbiol 1991 5 1531 1538 1787801
Ayala BP Vasquez B Clary S Tainer JA Rodland K So M The pilus-induced Ca2+ flux triggers lysosome exocytosis and increases the amount of Lamp1 accessible to Neisseria IgA1 protease Cell Microbiol 2001 3 265 275 11298650 10.1046/j.1462-5822.2001.00112.x
Kallstrom H Islam MS Berggren PO Jonsson AB Cell signaling by the type IV pili of pathogenic Neisseria J Biol Chem 1998 273 21777 21782 9705315 10.1074/jbc.273.34.21777
Merz AJ Enns CA So M Type IV pili of pathogenic neisseriae elicit cortical plaque formation in epithelial cells Mol Microbiol 1999 32 1316 1332 10383771 10.1046/j.1365-2958.1999.01459.x
Chen T Grunert F Medina-Marino A Gotschlich EC Several carcinoembryonic antigens (CD66) serve as receptors for gonococcal opacity proteins J Exp Med 1997 185 1557 1564 9151893 10.1084/jem.185.9.1557
Gray-Owen SD Dehio C Haude A Grunert F Meyer TF CD66 carcinoembryonic antigens mediate interactions between Opa- expressing Neisseria gonorrhoeae and human polymorphonuclear phagocytes. EMBO J 1997 16 3435 3445 9218786 10.1093/emboj/16.12.3435
Lee SW Bonnah RA Higashi DL Atkinson JP Milgram SL So M CD46 is phosphorylated at tyrosine 354 upon infection of epithelial cells by Neisseria gonorrhoeae. J Cell Biol 2002 156 951 957 11901164 10.1083/jcb.200109005
Howie HL Glogauer M So M The N. gonorrhoeae type IV pilus stimulates mechanosensitive pathways and cytoprotection through a pilT-dependent mechanism. PLoS Biol 2005 3 e100 15769184 10.1371/journal.pbio.0030100
Lee SW Higashi DL Snyder A Merz AJ Potter L So M PilT is required for PI(3,4,5)P3-mediated cross-talk between N. gonorrhoeae and epithelial cells. Cell Microbiol 2005 in press
Binnicker MJ Williams RD Apicella MA Gonococcal porin IB activates NF-kappaB in human urethral epithelium and increases the expression of host antiapoptotic factors. Infect Immun 2004 72 6408 6417 15501771 10.1128/IAI.72.11.6408-6417.2004
Roe BA Lin SP Song L Yuan X Clifton S Dyer D Gonococcal genome sequencing project. URL http://www.genome.ou.edu/gono/html 2002
Du Y Arvidson CG Identification of ZipA, an SRP-targeted protein from Neisseria gonorrhoeae. J Bacteriol 2003 185 2122 2130 12644481 10.1128/JB.185.7.2122-2130.2003
Du Y Lenz J Arvidson CG Global gene expression and the role of sigma factors in Neisseria gonorrhoeae in interactions with epithelial cells. Infect Immun 2005 73 4834 4845 16040997 10.1128/IAI.73.8.4834-4845.2005
Seifert HS Ajioka RS Paruchuri D Heffron F So M Shuttle mutagenesis of Neisseria gonorrhoeae: pilin null mutations lower DNA transformation competence J Bacteriol 1990 172 40 46 2152910
Goodman SD Scocca JJ Identification and arrangement of the DNA sequence recognized in specific transformation of Neisseria gonorrhoeae Proc Natl Acad Sci USA 1988 85 6982 6986 3137581
Dempsey JF Litaker W Hadhure A Snodgrass TL Cannon JG Physical map of the chromosome of Neisseria gonorrhoeae FA1090 with locations of genetic markers, including opa and pil genes. J Bacteriol 1991 173 5476 5486 1679431
Cotter PA Miller JF In vivo and ex vivo regulation of bacterial virulence gene expression. Curr Opin Microbiol 1998 1 17 26 10066465 10.1016/S1369-5274(98)80138-0
Dorman CJ Porter ME The Shigella flexneri virulence gene regulatory cascade: A paradigm of bacterial control mechanisms. Mol Microbiol 1998 29 677 684 9723908 10.1046/j.1365-2958.1998.00902.x
Lucas RL Lee CA Unravelling the mysteries of virulence gene regulation in Salmonella typhimurium. Mol Microbiol 2000 36 1024 1033 10844688 10.1046/j.1365-2958.2000.01961.x
Pettersson J Nordfelth R Dubinina E Bergman T Gustafsson M Magnusson KE Wolf-Watz H Modulation of virulence factor expression by pathogen target cell contact. Science 1996 273 1231 1233 8703058
Skorupski K Taylor RK Control of the ToxR virulence regulon in Vibrio cholerae by environmental stimuli. Mol Microbiol 1997 25 1003 1009 9350858 10.1046/j.1365-2958.1997.5481909.x
Sebastian S Agarwal S Murphy JR Genco CA The gonococcal fur regulon: identification of additional genes involved in major catabolic, recombination, and secretory pathways. J Bacteriol 2002 184 3965 3974 12081969 10.1128/JB.184.14.3965-3974.2002
West S Sparling F Response of Neisseria gonorrhoeae to iron limitation: alterations in expression of membrane proteins without apparent siderophore production. Infect Immun 1985 47 388 394 3155708
Hagman KE Shafer WM Transcriptional control of the mtr efflux system of Neisseria gonorrhoeae. J Bacteriol 1995 177 4162 4165 7608095
Lee EH Rouquette-Loughlin C Folster JP Shafer WM FarR regulates the farAB-encoded efflux pump of Neisseria gonorrhoeae via an MtrR regulatory mechanism. J Bacteriol 2003 185 7145 7152 14645274 10.1128/JB.185.24.7145-7152.2003
McKevitt M Patel K Smajs D Marsh M McLoughlin M Norris SJ Weinstock GM Palzkill T Systematic cloning of Treponema pallidum open reading frames for protein expression and antigen discovery. Genome Res 2003 13 1665 1674 12805273 10.1101/gr.288103
Nicholson TL Olinger L Chong K Schoolnik G Stephens RS Global stage-specific gene regulation during the developmental cycle of Chlamydia trachomatis. J Bacteriol 2003 185 3179 3189 12730178 10.1128/JB.185.10.3179-3189.2003
Shaw JH Falkow S Model for invasion of human tissue culture cells by Neisseria gonnorrhoeae. Infect Immun 1988 56 1625 1632 3131248
McGee ZA Johnson AP Taylor-Robinson D Human fallopian tubes in organ culture: preparation, maintenance, and quantitation of damage by pathogenic microorganisms. Infect Immun 1976 13 608 618 816745
McGee ZA Johnson AP Taylor-Robinson D Pathogenic mechanisms of Neisseria gonorrhoeae: observations on damage to human Fallopian tubes in organ culture by gonococci of colony type 1 or type 4. J Infect Dis 1981 143 413 422 6785363
McGee Z Stephens D Hoffman L Schlech III W Horn R Mechanisms of mucosal invasion by pathogenic Neisseria. Rev Infect Dis 1983 5 S708 S714 6415784
Jerse AE Experimental gonococcal genital tract infection and opacity protein expression in estradiol-treated mice. Infect Immun 1999 67 5699 5708 10531218
Rozen S Skaletsky H Krawetz SMS Primer3 on the WWW for general users and for biologist programmers. Bioinformatics Methods and Protocols Methods in Molecular Biology 2000 , Humana Press, Totowa, NJ 365 386
Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970 227 680 685 5432063 10.1038/227680a0
|
16137322
|
PMC1236931
|
CC BY
|
2021-01-04 16:38:05
|
no
|
BMC Microbiol. 2005 Sep 1; 5:50
|
utf-8
|
BMC Microbiol
| 2,005 |
10.1186/1471-2180-5-50
|
oa_comm
|
==== Front
BMC NeurolBMC Neurology1471-2377BioMed Central London 1471-2377-5-161616227810.1186/1471-2377-5-16Case ReportPrimary angiitis of the central nervous system presenting with subacute and fatal course of disease: a case report Lukas Carsten [email protected] Kathy [email protected]örnke Christian [email protected] Department of Neurology, St. Josef-Hospital, Ruhr-University Bochum, Gudrunstr. 56, 44791 Bochum, Germany2 Institute of Neuropathology, Westfälische Wilhelms-University, Domagkstr. 19, 48149 Münster, Germany2005 14 9 2005 5 16 16 6 6 2005 14 9 2005 Copyright © 2005 Lukas et al; licensee BioMed Central Ltd.2005Lukas 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
Primary angiitis of the central nervous system is an idiopathic disorder characterized by vasculitis within the dural confines. The clinical presentation shows a wide variation and the course and the duration of disease are heterogeneous. This rare but treatable disease provides a diagnostic challenge owing to the lack of pathognomonic tests and the necessity of a histological confirmation.
Case presentation
A 28-year-old patient presenting with headache and fluctuating signs of encephalopathy was treated on the assumption of viral meningoencephalitis. The course of the disease led to his death 10 days after hospital admission. Postmortem examination revealed primary angiitis of the central nervous system.
Conclusion
Primary angiitis of the central nervous system should always be taken into consideration when suspected infectious inflammation of the central nervous system does not respond to treatment adequately. In order to confirm the diagnosis with the consequence of a modified therapy angiography and combined leptomeningeal and brain biopsy should be considered immediately.
==== Body
Background
Primary angiitis of the central nervous system (PACNS) is an idiopathic, recurrent vasculitis confined to the central nervous system and the dural reflections. It may involve small and medium-sized leptomeningeal, cortical and subcortical arteries. The most common presentation is headache with encephalopathy accompanied by multifocal symptoms [1]. PACNS can occur at any age but there is a predominant appearance in the fourth to sixth decade [2]. The course and the duration of this disease show a wide variation. We describe a case with rapid progression and fatal outcome.
Case presentation
A 28-year-old man complaining of acute hoarseness and tingling in his left upper and lower extremities was admitted to hospital. His past medical history revealed persistent frontal headache for the last 3 weeks. On admission the patient was nervous and slightly confused. Fever was not present, although the patient complained of flu-like symptoms 4 weeks before. On neurologic examination dysarthria, a left facial weakness, clumsiness in rapid alternating movements of the left hand and dysmetric finger-to-nose testing beside abnormalities in pain and temperature sensation on the left side were found. Cranial computed tomography (CCT) scan was normal. Routine laboratory examinations showed no abnormalities. Serum antibodies to neurotropic viruses, HIV, Treponema pallidum and Borrelia burgdorferi were not found. The cerebrospinal fluid (CSF) contained inflammatory cells (32 lymphocytes/μl) and total CSF protein was elevated (930 mg/l). The glucose ratio was normal. Immunological studies and polymerase chain reactions for herpes simplex virus type 1, 2 and varicella-zoster virus were unobtrusive. Suspecting acute viral meningoencephalitis the patient was treated with aciclovir 3 × 750 mg/d. Within 24 hours after acute onset of focal neurologic symptoms remission of dysarthria was observed while the abnormal sensation and mild agitation were still present. Magnetic resonance (MR) scans performed 24 hours after acute onset of focal symptoms showed disseminated hyperintensities of the supratentorial grey and white matter with predominance in the right occipital region lacking enhancement after GdDTPA-injection (Figure 1). The MR-angiography was normal. The characteristics and location of these findings were not typical for acute disseminated encephalomyelitis or herpes simplex encephalitis. The follow-up MR scans showed an increase of white matter hyperintensities with enhancement after GdDTPA-injection. Repeated CSF analyses revealed no new aspects. During the hospital course only subfebrile body temperatures could be registered. Eight days after admission conventional angiography was considered but the patient developed severe disorientation and agitation. Within 12 hours he became unconscious and developed signs of transtentorial herniation. The emergency CCT-scan revealed massive generalized brain edema. Artificial ventilation was necessary. The patient was treated with steroids 2 × 1000 mg/d, but died 10 days after onset of his focal symptoms with signs of extensive brain stem dysfunction.
Figure 1 Cranial MRI (FLAIR) on the day of admission showing multiple lesions of the supratentorial grey and white matter.
Autopsy findings
The postmortem examination revealed primary angiitis of the central nervous system (PACNS) with focal granulomatous inflammation and partially thrombotic occlusion of leptomeningeal and cerebral arteries (Figure 2) resulting in multiple acute infarction areas in the cerebrum. There was also a brisk acute hypoxic-ischemic encephalopathy with diffuse eosinophilic neuronal injury (particularly in the cerebral cortex, hippocampus, Purkinje cell layer, and olivary nucleus) leading to and sustaining a prominent swelling of the brain with terminal herniation and necrosis of the parahippocampal gyrus as well as tonsillar tissue. There was no deposition of amyloid beta protein (Abeta) in the brain parenchyma and vessel walls as revealed by immunohistochemistry for Abeta (6F/3D anti-Aβ monoclonal antibody to residues 8 – 17, dilution1:100, Dako). Outside the dural confines no signs of vasculitis or systemic disease were found.
Figure 2 Leptomeningeal arteries show severe transmural inflammation with lumen obliteration (asterisk). The arrow indicates a fibrinoid necrosis in the vessel wall.
Conclusion
The clinical course of PACNS shows a wide variation, progressive courses with fatal outcome as well as benign courses have been reported. In general, the course of the untreated disease is, if clinical manifestation is protean, progressive, with death occurring within 9–12 months [2]. Shukla et al. reported on a patient with PACNS and an unusual prolonged clinical course of 8.5 years without treatment [3]. The ante-mortem diagnosis of PACNS depends on a combination of clinical, imaging and histological findings. Major differential diagnosis of PACNS include secondary vasculitides of the CNS which can occur in collagen vascular disorders, viral and bacterial infections, in malignancies and due to drugs and substance abuse [1,4]. Abeta-related angiitis (ABRA) that occasionally complicates cerebral amyloid angiopathy may also be taken into consideration, however, the mean age of presentation of ABRA seems to be higher than that of PACNS [5]. Young women with a history of headaches may present with benign angiopathy of the central nervous system [6,7]. Headache and encephalopathy are also common clinical features of posterior leukoencephalopathy. This newly recognised neurological disorder is usually seen in patients with uncontrolled hypertension, eclampsia and in patients treated with immunosuppressive drugs [8]. As recently published, posterior leukoencephalopathy can be associated with isolated CNS angiitis [9].
Examination of the CSF in PACNS is abnormal in 80–90% of patients and consists of lymphocytic pleocytosis and elevated protein [6]. MRI findings in PACNS are variable and non-specific. The most common findings are multiple bilateral asymmetrical supratentorial lesions, involving grey and white matter [10]. In a few cases mass lesions have been described [10,11], but nevertheless normal MRI and biopsy findings consistent with CNS vasculitis have also been found [12]. Abnormal conventional angiography may be useful to assume cerebral vasculitis [12,13], but many cases of PACNS have normal arteriograms while in cases with abnormal findings, such as multiple segmental narrowings of cerebral arteries, the diagnosis of PACNS could not be confirmed [14]. Moore has suggested criteria for the diagnosis of PACNS [2]. To fulfill these criteria systemic infections and inflammatory diseases have to be excluded. Combined leptomeningeal and brain biopsy has to confirm the presence of vascular inflammation. Biopsy should be performed prior to immunosuppressive treatment in order to exclude lesions of infectious, neoplastic or non-inflammatory etiology and to confirm a diagnosis of vasculitis. Due to the segmental nature of the lesions false-negative results can be obtained in up to 35% of biopsies [15-18]. The best established therapy is a combination of high-dose steroids with cyclophosphamide [19]. After clinical stabilization of the patient, treatment should be continued for at least 1 year [20].
In our case the duration between onset of neurological symptoms and death was markedly short. Fever, that one would expect in infectious inflammation of the CNS, was not present in our patient. Nevertheless, the lack of fever is not specific enough to guide diagnostic efforts. These conditions highlight the importance of early angiography and biopsy in cases of suspected vasculitis. Unfortunately, in our case the brief and fluctuating course of the disease with final deterioration within 10 days did not lead to angiography and brain biopsy in time. In conclusion even in young patients PACNS should always be taken into consideration especially when suspected infectious inflammation of the CNS does not respond to treatment adequately. In the face of a progressive clinical course and no specific etiological diagnosis one should not continue to treat CNS inflammatory disease empirically. Since PACNS would require a modification of the therapy, angiography and a combined leptomeningeal and cortical biopsy should be performed immediately.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CL carried out the neurological examination, study of literature and substantial contributions in writing and design of the manuscript. CL prepared Figure 1. KK carried out the neuropathological investigations, description of this investigation and prepared Figure 2. CB reviewed and corrected the manuscript and gave final approval of the version to be published.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Written consent was obtained from the relatives of the patient for publication of the case report.
==== Refs
Siva A Vasculitis of the nervous system J Neurol 2001 248 451 468 11499635 10.1007/s004150170154
Moore PM Richardson B Neurology of the vasculitides and connective tissue diseases J Neurol Neurosurg Psychiatry 1998 65 10 22 9667555
Shukla G Deol PS Arora R Prasad K Behari M Isolated angiitis of the central nervous system: report of a patient with an unusually prolonged course Eur Neurol 2001 46 162 163 11598339 10.1159/000050794
Rollnik JD Brandis A Dehghani K Bufler J Lorenz M Heidenreich F Donnerstag F [Primary angiitis of CNS (PACNS)] Nervenarzt 2001 72 798 801 11688183 10.1007/s001150170038
Scolding NJ Joseph F Kirby PA Mazanti I Gray F Mikol J Ellison D Hilton DA Williams TL MacKenzie JM Xuereb JH Love S Abeta-related angiitis: primary angiitis of the central nervous system associated with cerebral amyloid angiopathy Brain 2005 128 500 15. Epub 2005 Jan 19. 15659428 10.1093/brain/awh379
Calabrese LH Duna GF Lie JT Vasculitis in the central nervous system Arthritis Rheum 1997 40 1189 1201 9214418
Hajj-Ali RA Furlan A Abou-Chebel A Calabrese LH Benign angiopathy of the central nervous system: cohort of 16 patients with clinical course and long-term followup Arthritis Rheum 2002 47 662 669 12522842 10.1002/art.10797
Hinchey J Chaves C Appignani B Breen J Pao L Wang A Pessin MS Lamy C Mas JL Caplan LR A Reversible Posterior Leukoencephalopathy Syndrome N Engl J Med 1996 334 494 500 8559202 10.1056/NEJM199602223340803
Wijdicks EF Manno EM Fulgham JR Giannini C Cerebral angiitis mimicking posterior leukoencephalopathy J Neurol 2003 250 444 448 12700910 10.1007/s00415-003-1021-4
Singh S John S Joseph TP Soloman T Primary angiitis of the central nervous system: MRI features and clinical presentation Australas Radiol 2003 47 127 134 12780440 10.1046/j.0004-8461.2003.01140.x
Ceccarelli A De Blasi R Pavone I Lamberti P Carella A Livrea P Simone IL Primary angiitis of the central nervous system: a misinterpreted clinical onset of CNS vasculitis Eur Neurol 2005 53 40 42 15746543 10.1159/000084263
Alhalabi M Moore PM Serial angiography in isolated angiitis of the central nervous system Neurology 1994 44 1221 1226 8035919
Campi A Benndorf G Filippi M Reganati P Martinelli V Terreni MR Primary angiitis of the central nervous system: serial MRI of brain and spinal cord Neuroradiology 2001 43 599 607 11548164 10.1007/s002340100561
Kadkhodayan Y Alreshaid A Moran CJ Cross DT Powers WJ Derdeyn CP Primary angiitis of the central nervous system at conventional angiography Radiology 2004 233 878 82. Epub 2004 Oct 21. 15498898
Chu CT Gray L Goldstein LB Hulette CM Diagnosis of intracranial vasculitis: a multi-disciplinary approach J Neuropathol Exp Neurol 1998 57 30 38 9600195
Duna GF Calabrese LH Limitations of invasive modalities in the diagnosis of primary angiitis of the central nervous system J Rheumatol 1995 22 662 667 7791160
Lie JT Primary (granulomatous) angiitis of the central nervous system: a clinicopathologic analysis of 15 new cases and a review of the literature Hum Pathol 1992 23 164 171 1740300 10.1016/0046-8177(92)90238-X
Alrawi A Trobe JD Blaivas M Musch DC Brain biopsy in primary angiitis of the central nervous system Neurology 1999 53 858 860 10489055
Calabrese LH Therapy of systemic vasculitis Neurol Clin 1997 15 973 991 9367976 10.1016/S0733-8619(05)70359-5
Riemer G Lamszus K Zschaber R Freitag HJ Eggers C Pfeiffer G Isolated angiitis of the central nervous system: lack of inflammation after long-term treatment Neurology 1999 52 196 199 9921877
|
16162278
|
PMC1236932
|
CC BY
|
2021-01-04 16:28:53
|
no
|
BMC Neurol. 2005 Sep 14; 5:16
|
utf-8
|
BMC Neurol
| 2,005 |
10.1186/1471-2377-5-16
|
oa_comm
|
==== Front
BMC NeurosciBMC Neuroscience1471-2202BioMed Central London 1471-2202-6-571613525210.1186/1471-2202-6-57Research ArticleNeuromagnetic brain responses to words from semantic sub- and supercategories Assadollahi Ramin [email protected] Brigitte [email protected] Department of Psychology, University of Konstanz, P.O. Box D23 D-78457 Konstanz, Germany2005 31 8 2005 6 57 57 27 10 2004 31 8 2005 Copyright © 2005 Assadollahi and Rockstroh; 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
We explored spatio-temporal patterns of cortical activity evoked by written words from super-ordinate and sub-ordinate semantic categories and hoped to find a differential cortical and/or temporal distribution of the brain response depending on the level of the categories. Twenty-three subjects saw 360 words belonging to six sub-ordinate categories (mammals, birds, fish, fruit, flowers, trees) within two super-ordinate categories (fauna, flora). Visually evoked magnetic fields were determined from whole-head (148-sensor) magnetoencephalography and analyzed in the source space (Minimum Norm Estimate).
Results
Activity (MNE amplitudes) 100–150 ms after stimulus onset in the left occipito-temporal area distinguished super-ordinate categories, while later activity (300–550 ms) in the left temporal area distinguished the six sub-ordinate categories.
Conclusion
Our results document temporally and spatially distinct processing and representation of words according to their categorical information. If further studies can rule out possible confounds then our results may help constructing a theory about the internal structure of entries in the mental lexicon and its access.
==== Body
Background
The mental lexicon is usually considered to be a compilation of linguistic information, with each word having an entry. The precise structure of a lexical entry continues to be a topic of investigation and discussion. From a linguistic perspective, a lexical entry for a word should contain information about its pronunciation, its orthography, its syntactic properties, and its semantic information. In the present study, we explored the aspect of categories within semantic information. A sparrow for example, might belong to the category bird as well as to the category animal. How and where is such information encoded in the brain? Is the category information represented within one brain compartment or in a distributed fashion?
These questions are of course not only restricted to the mental lexicon but also apply to the field of visual object recognition, for example. Cognitive neuroscience has contributed to these questions by considering evidence on the representation of object categories in the brain (mainly derived from picture naming tasks): For visually presented exemplars of particular object categories the use of functional brain imaging disclosed activity in the ventral pathway of the visual system, with activity distinguishing different object categories (faces, animals, fruits, buildings, chairs, and man made tools) along the occipito-temporal pathway, which was thereupon labelled 'what-stream' (e.g., [1-9]).
Evidence from brain responses distinguishing between sub- and super-ordinate category information is still limited. For instance, Löw and colleagues [10] examined the spatial distribution of sources of neuromagnetic activity elicited by pictures of super- (e.g. animal, furniture etc) and sub-ordinate (e.g., deer, wolf; table, chair, etc.) object categories. Significantly higher similarity (correlation) of evoked magnetic responses within than across super-ordinate categories indicated macroscopically distinct representation of defined semantic categories. Activation of these representations was most prominent in the right parietal cortex around 200 ms and in the left parieto-temporal cortex around 400 ms after stimulus onset. This spatio-temporal sequence supports sequential perceptual dynamics of the categorization process, with initial activation of representation in the extrastriate areas of the right hemisphere followed by activity along the ventral processing stream primarily in the left temporal lobe.
This evidence served as the background for the present study, which aimed at clarifying the dynamics involved in the lexical access of words. Representation and activation of semantic categories were probed by analyzing spatio-temporal brain activity that was induced by written words. They a) were related to different super- and sub-ordinate categories; b) came from the natural domain (in order to rule out/control for possible effects of gross features of the man-made/nature-made difference); c) were of low to ultra-low frequency (exploiting evidence that low-frequent words induce marked event-related brain responses, [11]). Magnetic source imaging (MSI, the analysis of sources of activation from high-resolution magnetoencephalography) was used as an indication of cortical representation. In particular, we aimed at clarifying the following questions: a) whether visually presented words related to different super- and sub-ordinate natural categories would be represented in spatially separable areas along the occipital-to-temporal visual pathway, corresponding to the 'what stream' determined for visually presented object categories; b) whether distinct activity patterns primarily in the left hemisphere would indicate the representation of (additional) linguistic features for semantic categories activated by words; c) whether spatio-temporal brain responses distinguish super- and sub-ordinate categories as an indication of feature-directed category representation and access of lexical entries.
Results and discussion
Highly similar activity patterns were elicited by all six sub-ordinate categories and we did not observe gross topographical differences in response to different categories. Therefore, Figure 1 illustrates the course of activity analyzed in the source space and averaged across the six categories. A left occipital focus of activity is prominent already for the first interval 100–150 ms after stimulus-onset and remains obvious throughout the recording period; starting around 250 and 300 ms, an additional left-temporal focus becomes evident (see arrow in Fig. 1); as a third prominent focus, left-anterior activation starting around 350 ms can be taken form Fig. 1.
As is also obvious from Figure 1, activity was prominent in the left hemisphere and negligible in the right hemisphere. Since effects of stimulus categories on activity patterns were statistically confirmed only for the left but not for the right hemisphere, the subsequent results refer to the left-hemisphere foci.
Figure 2a illustrates the time course of source activity (amplitudes of Minimum Norm Estimates) at the left occipital area (covering 7 dipoles) separately for the two super-ordinate categories. More pronounced activity induced by fauna- compared to flora-items is indicated around 150 ms after stimulus-onset, which is confirmed for the time segment in the difference curve (Figure 2b) and by a main effect SUPERCATEGORY (comparing the super-ordinate categories; F(1,22) = 4.7, p < .05). Differences in subsequent time segments fell short of significance. The source localization of this effect may suggest that this activity is related to processing of the visual word form in the so-called Visual Word Form Area (VWFA, c.f. Salmelin group: [12], Dehaene group: [13]). The comparison with results from our own study [14] using the same localization methodology as in the present study revealed that the two areas are in fact distinct. The focus of activity appears to be more anterior and more superior compared to the focus reported in [14]. This suggests that the activation difference in the SUPERCATEGORY is not due to simple visual word form processing (and possibly related factors like number of vowels, morphemes, syllables, word frequency and the like). Studies finding differences in the VWFA, usually modulated low level properties of the stimulus material such as words vs. non-words vs. pseudo-words vs. symbols. Our own study [14] manipulated word length and frequency. Whereas the modulation of length revealed different amplitudes in the occipital lobe, the differential activation due to variation of word frequency was localized at the VWFA. We are not aware of any study that found different activations due to semantic variation of the stimulus material. Thus, effects of wordness and semantics may occur in similar time ranges but at different areas in the brain.
The time course of activity for the left temporal region (Figure 3) discloses differences between sub-ordinate categories between 300 and 550 ms after stimulus onset. Interactions SUPER × SUBCATEGORY indicate significant differences between sub- and super-ordinate categories for the intervals 300–350 ms (F(2,44) = 3.3, p < .05), 350–400 ms (F(2,44) = 3.1; p = .05), and 400–450 ms (F(2,44) = 3.8, p < .05). For the last interval post-hoc tests verified significant differences between the sub-categories fruit vs flowers (p < .05), fruit vs. trees (p < .01) and fruit vs. fish (p < .01) and marginally significant trends (p < .1) for differences between birds vs fruit, birds vs trees, and mammals vs fruit. Distinct category processing remained apparent until 550 ms after stimulus onset, with the difference for 450–500 ms approaching significance (p < .1) and the difference between 500 and 550 ms (F(2,44) = 4.0; p < .05) being explained by significant post-hoc comparisons of fruit vs. flowers (p < .01), fruit vs. trees (p < .01) and fruit vs. fish (p < .05), flowers vs birds (p < .05) and flowers vs mammals (p < .05), trees vs birds (p < .01) and trees vs mammals (p < .01), and fish vs mammals (p = .05).
In sum, the activity near or at the left temporal pole differentiates between sub-categories. The sub-categories were selected to be semantic in nature. It is not clear whether the activity found resembles an N400 (a component generally reflecting the ease at which a word is integrated into a sentential context), which was also found to be elicited after single words [15]. Also, source localisations of the N400 component do not give a clear picture where the N400 has its cortical origin.
For the left-anterior activity focus indicated in Figure 1, no significant effects of categories were found.
In sum, source analyses of neuromagnetic brain responses to words from different semantic categories, comprising super-ordinate (fauna, flora) and sub-ordinate categorization (mammals, birds, fish; fruit, flowers, trees) disclose specific spatio-temporal activation and a distinction in time and space for sub- versus super-ordinate category features: Whereas super-ordinate features induced different activation already around 100–150 ms after stimulus onset in left-occipital areas, the processing of sub-ordinate category features led to differential activation only later, between 300 and 550 ms, in left-temporal areas. This is, to our knowledge, the first evidence of spatio-temporal separation of category-feature processing in the brain.
Topography
"Hebbian-type cell assemblies" are described as distributed networks of neurons that are functionally connected by reciprocal dynamic connections and whose strength is modulated based on correlation learning [16,17]. In Pulvermüller's framework of the mental lexicon (see [16] for an overview), the neuronal representation of a word may comprise neurons in the perisylvian cortex (where the acoustic word form is represented) and neurons in areas that process information about object features or associated actions. For nouns denominating objects these features comprise visual associations and, thus, the neuronal network activated by the noun will comprise neurons in the visual association areas in the occipital and inferior temporal lobe. During language acquisition, representations of objects are related to auditory or visual word forms. A cell assembly is established whenever a concept is activated repeatedly and simultaneously with the word form. This cell assembly will then comprise features that reflect parts of its word form representation and parts of the auditory or visual representation of the object the word refers to (the so-called referential information). From this point of view it makes sense that semantic features of object representations are also shared by the corresponding entries in the mental lexicon. Thus, the same sub-assemblies (reflecting a certain feature) are activated either by the word or by the object itself.
Pulvermüller's work also relates to the research field of the ventral "what-stream". Ishai [18] pointed out that the topological arrangement of visual features in the ventral "what"-stream remains to be clarified. If, as Pulvermüller argues, category-features activated by words are part of category-features activated by corresponding objects, then the present results contribute to this clarification, in that features related to sub-ordinate concepts are processed in areas more anterior to features related to super-ordinate concepts. Gauthier and colleagues [19] pointed out that in visual object recognition the super-ordinate-level (e.g. 'bird'), at which familiar objects are first recognized, was distinguished from more sub-ordinate levels (e.g. 'sparrow'), which required additional perceptual processing. This additional processing related to the sub-ordinate level representation was associated with activation in the fusiform and inferior temporal gyri and the temporal poles. This is in line with the more anterior focus of activity distinguishing sub-ordinate categories in the present study.
In the present study, it is conceivable that the categories fauna and flora activated features related to the visual representations of the respective objects. These features would be part of the representations of the physical, visually perceivable objects and as such part of the ventral "what"-stream of the left occipito-temporal cortex. There are several studies that showed early activation due to referential semantic information (e.g. [20-23]).
While these considerations are supported by the present results of category-specific activity patterns along the left occipito-temporal axis, they should be validated by a direct comparison of cortical activity evoked by pictures and their corresponding words. Chao and colleagues [5] report an fMRI study, where in fact there were similar activations for pictures and their corresponding names.
Why should the representations of sub- and supercategory features be processed at different sites in the brain? There is ample evidence that during the individual development of the brain, maximum neural plasticity starts at primary sensory (e.g. occipital) and motor sites and over time progresses towards secondary areas, parietal areas and finally towards frontal areas [24-28]. Using computer simulations, Shrager and Johnson [29] showed that simple representations of input data can be acquired in early stages of development (i.e. plasticity at primary sites) whereas more abstract representations can only be acquired in sites that had high plasticity later on (i.e. more anterior sites). These sites would use the information from the lower level representations together with the input to represent more abstract and/or invariant aspects of the stimulus. In the present case, this would mean that gross categories can be represented early on (in more primary, posterior sites) whereas more fine grained sub-representations will only be acquired later on (in more anterior sites).
Amplitude
We can only speculate how the amplitude difference in the present study relates to different categories or their features. Generally, there are two reasons for a stronger amplitude in the MEG. Either there are slight topographic differences or the size of the neuronal network activated varies over conditions. Cortical sources may be located on a part of a gyrus where the tangential part of the magnetic field is more or less prevalent than in a different condition. The magnetometers used in our 4D Neuroimaging system only capture tangential fields and thus a stronger component in the radial field would mean a reduction in the field strength measured. Thus, the amplitude difference may arise from a slight topographic shift. The other possible answer is that different features have different complexities in their representation. This may translate in different sizes of neuronal networks representing them. The activation of these networks will lead to greater amplitudes for larger networks. Another reason for different network sizes may be the relation of the word to other words. If all words are represented in networks, then these networks may have overlapping parts to reflect the relatedness of words. We tried to capture the associations that subjects had after reading our stimuli and found no systematic variation in the number of associations over categories. Thus, if networks of words (i.e. networks of networks) lead to different amplitudes due to a different degree of connectedness, then this is not a good explanation for the amplitude differences found in the present study. We therefore favour the interpretation of slightly different topographies resulting in different amplitudes.
Timeline
The present study not only revealed that category related information is distributed spatially over the cortex, but it is also activated over time. The temporal sequence of cortical activation distinguished the processing super-ordinate (earlier) and sub-ordinate features (later). Considering the results of Gauthier and colleagues [19], we may conclude that features of lexical entries are hierarchically processed, with steps of feature analyses (super-ordinate first) spreading across hundreds of ms.
Conclusion
The present results may help to speculate over the inner structure of entries in the mental lexicon: Super- and sub-ordinate category features are parts of a word's representation; they activate distinct cortical areas in the left occipital-temporal axis at distinct points in time. Lexical entries may thus be understood as structure of features, represented in distributed networks that are activated by words in a sequential, cascaded way.
Methods
Subjects
Twenty-three healthy, right-handed, native German speaking student subjects (12 m, 11f, mean age: 26) volunteers with normal or corrected-to-normal vision were paid (15€ for the two-hour session) for participation in the study.
Materials & procedure
Stimulus selection
Stimuli comprised 360 words related to the two super-ordinate categories (labelled 'fauna' and 'flora') and 6 sub-ordinate categories (fauna: mammals, birds, fish; flora: fruit, flowers, trees). For each sub-ordinate category, 60 stimuli were selected from the CELEX-database [30] for low-frequency (mean frequency was seven per million) and comparable length (8 letters). Since CELEX did not provide enough low frequent nature-made object names meeting these criteria (match for length, frequency < 10), 10 words with ultra-low frequency were added from other sources to each sub-category. We wanted to have as many stimuli as possible to have a broad base of individual concepts sharing their sub- and supercategory features. A better match for other word-level features such as bigram-letter-frequency etc would have further reduced the number of stimuli per category and would thus have reduced the representativeness of the study. The 360 words were assigned to 2 runs, each of 6 series of 30 stimuli (per sub-ordinate category). Within each series the 30 stimuli were presented in pseudo-random order.
Behavioural study
Using a computerized test, we presented all 360 stimuli to 48 subjects (30 randomly selected items per subject; subjects did not take part in the MEG study) and asked them to write up their associations. Subjects had 60 seconds time to type in their association for each stimulus. The procedure resulted in four (= 360/48) sets of associations per stimulus, 240 (= 60 × 4) sets of associations per sub-ordinate category and 720 (= 240 × 3) sets of associations per super-ordinate category. We investigated the number of associations per word (i.e. how many associations were made on the average for this particular word) using a computer program and were surprised that this number was surprisingly similar for the super-ordinate categories (fauna: mean = 7.217, SD = 7.463; flora: mean = 7.206, SD = 7.414) as well as for the sub-ordinate categories (birds: mean = 7.289, SD = 7.499; mammals: mean = 7.347, SD = 7.582 ; fish: mean = 7.016, SD = 7.308; fruit: mean = 8.0, SD = 8.162; flowers: mean = 6.950, SD = 7.250; trees: mean = 7.206, SD = 7.414). Also, there were no differences between words that appear in the CELEX database and words that do not.
Procedure
Stimuli were presented in white upper case letters (maximum word size 9 × 3 cm) on a black background at a distance of 1.4 m using an LCD-projector situated outside the MEG chamber. Presentation time was 250 ms per word with inter-stimulus intervals varying between 1200 and 2000 ms and an average of 1600 ms.
Subjects were instructed to fixate a cross which was projected on the screen during stimulus-free intervals, to read each word and to respond by button press only to (rare) nouns depicting artificial objects. These 36 stimuli were interspersed with 10% probability in order to ensure sustained attention and processing involving/activating the mental lexicon. (Responses to these filler items were not analysed.)
Data collection and analyses
Neuromagnetic signals were recorded continuously with a 148 channel whole head magnetometer (Magnes 2500 WH, 4D NeuroImaging Inc., San Diego) using a 0.1–100 Hz band-pass filter and sampled at a rate of 508 Hz. Additionally, vertical and horizontal EOG (electro-oculogram) were recorded for artifact control. Written informed consent was obtained from subjects prior to each MEG-session and the study was approved by the ethics committee of the University of Konstanz.
After external global noise subtraction continuous MEG and exclusion of artifact-contaminated epochs by visual inspection (EOG level > 100 μV, an MEG level > 5 pT, button press-induced artifacts), data were split into epochs of 900 ms length (including 100 ms pre-stimulus). This resulted on average in 50.2 MEG-traces per sub-category, subject and run (= 100.4 MEG-traces in sum per sub-category and subject). For every subject and for each of the six sub-ordinate categories, evoked magnetic fields (EMFs) were determined relative to a baseline of 100 ms pre-stimulus onset. For these average EMFs, cortical sources were determined using the minimum norm estimate (MNE, [31-33]), an objective inverse method to reconstruct the topography of the primary current underlying a magnetic field distribution ([34]). Cortical activity was approximated in a three-dimensional spherical source space of radius 10 cm fitted individually to the head-shape of the subjects (4-D Neuroimaging software). On this sphere 197 equidistant dipoles for the estimation of source activity were assumed. The source estimations (MNE) of the two runs per subcategory were averaged.
MNE amplitudes (in nAm) were averaged for 50 ms-segments across the 600-ms interval after stimulus onset and plotted separately for the two super- and six sub-ordinate categories, respectively. Since topographies for all six sub-categories were similar, time intervals and areas of interest for further analyses were determined for the average across all sub-categories. As apparent from Figure 1 foci of activity, that is, topographic peaks of MNE amplitudes, were obvious for three distinct areas in the left hemisphere. For these areas, the MNE amplitudes of seven dipoles were averaged for statistical analyses of stimulus effects. For each of the three areas (left-occipital, left-temporal and left-anterior), effects of categories on MNE-amplitudes were verified by two-way repeated measures Analyses of Variance comparing the three sub-ordinate categories each (as within-subject factor SUB-ORDINATE) within the two super-ordinate categories (as within-subject factor SUPER-ORDINATE) for successive 50-ms time intervals. Stars in figure 2 resemble significant main effects whereas stars in figure 3 resemble significant interactions. Newman-Keuls tests served for post-hoc explanation of interactions.
Authors' contributions
RA designed the study, collected the data and accomplished the data analyses, and drafted the manuscript. BR was involved in planning and designing the study and substantially contributed to the manuscript.
Acknowledgements
Research was supported by the DFG (German Research Community; SFB 471).
Figures and Tables
Figure 1 Topographies. Topographical distribution of source activity (MNE amplitudes) across 50-ms intervals in the left (upper rows) and right (lower rows) hemisphere, averaged across all stimuli and subjects. Darker areas indicate stronger source activation (MNE-amplitudes).
Figure 2 Source waveforms and difference waves for left occipital area. (a) Source waveform for left occipital area. Time course of activity (amplitudes of MNE in nAm) in the left-occipital area across 100 ms pre- and 600 ms post-stimulus onset averaged across subjects separately for the super-ordinate categories (fauna: bold, flora: dashed). (b) Difference waves for left occipital area. Time course of the difference in brain activity (amplitudes of MNE in nAm in the left-occipital area) between super-ordinate categories (fauna – flora). The time interval with the significant difference is marked by asterisks, an illustrative difference map is given as insert.
Figure 3 Source waveforms for left temporal area. Time course of brain activity (amplitudes of MNE in nAm) for 100 ms pre- and 600 ms post-stimulus onset averaged across subjects separately for the six sub-ordinate categories in the left-temporal area. Significant interactions are marked by asterisks, the marginally significant interaction by a tilde.
==== Refs
Martin A Wiggs CL Ungerleider LG Haxby JV Neural correlates of category-specific knowledge Nature 1996 379 649 652 8628399 10.1038/379649a0
Schendan HE Ganis G Kutas M Neurophysiological evidence for visual perceptual categorization of words and faces within 150 ms Psychophysiology 1998 35 240 251 9564744 10.1017/S004857729897010X
Aguirre GK Zarahn E D'Esposito M An area within human ventral cortex sensitive to 'building' stimuli: evidence and implications Neuron 1998 21 373 383 9728918 10.1016/S0896-6273(00)80546-2
Gerlach C Law I Gade A Paulson OB Perceptual differentiation and category effects in normal object recognition Brain 1999 122 2159 2170 10545400 10.1093/brain/122.11.2159
Chao LL Haxby JV Martin A Attribute-based neural substrates in temporal cortex for perceiving and knowing about objects Nat Neurosci 1999 2 913 919 10491613 10.1038/13217
Ishai A Ungerleider LG Martin A Schouten HL Haxby JV Distributed representation of objects in the human ventral visual pathway Proc Natl Acad Sci U S A 1999 96 9379 9384 10430951 10.1073/pnas.96.16.9379
Moore CJ Price CJ A functional neuroimaging study of the variables that generate category-specific object processing differences Brain 1999 122 943 962 10355678 10.1093/brain/122.5.943
Allison T Puce A Spencer DD McCarthy G Electrophysiological studies of human face perception. I: Potentials generated in occipitotemporal cortex by face and non-face stimuli Cereb Cortex 1999 9 415 430 10450888 10.1093/cercor/9.5.415
Kanwisher N Domain specificity in face perception Nat Neurosci 2000 3 759 763 10903567 10.1038/77664
Löw A Elbert T Bentin S Silberman Y Gomolla A Cohen R Rockstroh B Semantic categories reflected in macroscopically distinct human brain activities Psychol Sci 2003 14 367 372 12807412 10.1111/1467-9280.24451
Rugg MD Event-related brain potentials dissociate repetition effects of high- and low-frequency words Memory & Cognition 1990 18 367 379 2381316
Tarkiainen A Helenius P Hansen PC Cornelissen PL Salmelin R Dynamics of letter string perception in the human occipitotemporal cortex Brain 1999 122 2119 2132 10545397 10.1093/brain/122.11.2119
Cohen L Lehéricy S Chochon F Lemer C Rivaud S Dehaene S Language-specific tuning of visual cortex? Functional properties of the Visual Word Form Area Brain 2002 125 1054 1069 11960895 10.1093/brain/awf094
Assadollahi R Pulvermuller F Early influences of word length and frequency: a group study using MEG Neuroreport 2003 14 1183 1187 12821805 10.1097/00001756-200306110-00016
Bentin S McCarthy G Wood CC Event-related potential, lexical decision and semantic processing Electroencephalogr Clin Neurophysiol 1985 60 342 355 2579801
Pulvermüller F Words in the brain's language Behav Brain Sci 1999 22 253 336 11301524 10.1017/S0140525X9900182X
Varela FJ Lachaux JP Rodriguez E Martinerie J The Brain Web: Phase Synchronisation and Large-Scale Integration Nat Rev 2001 2 229 239 10.1038/35067550
Ishai A On Representation and Reproducibility J Cogn Neurosci 2003 15 946 947 14614805 10.1162/089892903770007335
Gauthier I Anderson AW Tarr MJ Skudlarski P Gore JC Levels of categorization in visual object studied with functional MRI Curr Biol 1997 7 645 651 9285718 10.1016/S0960-9822(06)00291-0
Marslen-Wilson WD Tyler LK The temporal structure of spoken language understanding Cognition 1980 8 1 71 7363578 10.1016/0010-0277(80)90015-3
Skrandies W Evoked potential correlates of semantic meaning : a brain mapping study Brain Res Cogn Brain Res 1998 6 173 183 9479069 10.1016/S0926-6410(97)00033-5
Perani D Cappa SF Schnur T Tettamanti M Collina S Rosa M Fazio F The neural correlates of verb and noun processing: a PET study Brain 1999 122 2337 2344 10581226 10.1093/brain/122.12.2337
Hauk O Pulvermüller F Neurophysiological Distinction of Action Words in the Fronto-Central Cortex Hum Brain Mapp 2004 21 191 201 14755838 10.1002/hbm.10157
Pandya DN Yeterian EH Scheibel AB, Wechsler AF Architecture and connections of cerebral cortex: implications for brain evolution and function The Neurobiology of higher cognitive function 1990 New York: Guilford Press 53 84
Thatcher RW Cyclic cortical reorganization during early childhood Brain Cogn 1992 20 24 50 1389121 10.1016/0278-2626(92)90060-Y
Chugani HT Phelps ME Mazziotta JC Positron emission tomography study of human brain functional development Ann Neurol 1987 22 487 497 3501693 10.1002/ana.410220408
Hanverth RS Smith ELI Duncan GC Crawford MLI von Noorden GK Multiple sensitive periods in the development of the primate visual system Science 1986 232 235 238 3952507
Shrager J Johnson ME Dynamic plasticity influences the emergence of function in a simple cortical array Neur Netw 1996 9 1119 1129 10.1016/0893-6080(96)00033-0
Baayan H Piepenbrock R van Rijn H The CELEX lexical database (CD-Rom) Linguistic Data Consortium 1993 University of Pennsylvania PA
Hämäläinen MS Hari R Ilmoniemi RJ Knuutila J Lounasmaa OV Magnetoencephalography – theory, instrumentation, and applications to noninvasive studies of the working human brain Reviews of Modern Physics 1993 65 413 497 10.1103/RevModPhys.65.413
Hämäläinen MS Ilmoniemi RJ Interpreting magnetic fields of the brain: minimum norm estimates Med Biol Eng Comput 1994 32 35 42 8182960
Hauk O Berg P Wienbruch C Rockstroh B Elbert T Yoshimoto T, Kotani M, Kuriki S, Karibe H, Nakasato N Recent advances in biomagnetism 1999 Sendai: Tohoku University Press 213 216
Grave de Peralta Menendez R Hauk O Gonzalez Andino S Vogt H Michel C Linear inverse solutions with optimal resolution kernels applied to the electromagnetic tomography Hum Brain Mapp 1997 5 454 467 10.1002/(SICI)1097-0193(1997)5:6<454::AID-HBM6>3.0.CO;2-2
|
16135252
|
PMC1236933
|
CC BY
|
2021-01-04 16:03:47
|
no
|
BMC Neurosci. 2005 Aug 31; 6:57
|
utf-8
|
BMC Neurosci
| 2,005 |
10.1186/1471-2202-6-57
|
oa_comm
|
==== Front
BMC PediatrBMC Pediatrics1471-2431BioMed Central London 1471-2431-5-371615329710.1186/1471-2431-5-37Research ArticleComparison of air displacement plethysmography to hydrostatic weighing for estimating total body density in children Claros Geo [email protected] Holly R [email protected] David A [email protected] Department of Health and Exercise Science, University of Oklahoma, Norman, OK, USA2 Department of Pediatrics, University of Oklahoma Health Science Center, Oklahoma City, OK, USA3 Children's Medical Research Institute's Metabolic Research Center, University of Oklahoma Health Science Center, OUCP Diabetes and Endocrinology, Oklahoma City, OK, USA2005 9 9 2005 5 37 37 18 4 2005 9 9 2005 Copyright © 2005 Claros et al; licensee BioMed Central Ltd.2005Claros 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 purpose of this study was to examine the accuracy of total body density and percent body fat (% fat) using air displacement plethysmography (ADP) and hydrostatic weighing (HW) in children.
Methods
Sixty-six male and female subjects (40 males: 12.4 ± 1.3 yrs, 47.4 ± 14.8 kg, 155.4 ± 11.9 cm, 19.3 ± 4.1 kg/m2; 26 females: 12.0 ± 1.9 yrs, 41.4 ± 7.7 kg, 152.1 ± 8.9 cm, 17.7 ± 1.7 kg/m2) were tested using ADP and HW with ADP always preceding HW. Accuracy, precision, and bias were examined in ADP with HW serving as the criterion method. Lohman's equations that are child specific for age and gender were used to convert body density to % fat. Regression analysis determined the accuracy of ADP and potential bias between ADP and HW using Bland-Altman analysis.
Results
For the entire group (Y = 0.835x + 0.171, R2 = 0.84, SEE = 0.007 g/cm3) and for the males (Y = 0.837x + 0.174, R2 = 0.90, SEE = 0.006 g/cm3) the regression between total body density by HW and by ADP significantly deviated from the line of identity. However in females, the regression between total body density by HW and ADP did not significantly deviate from the line of identity (Y = 0.750x + 0.258, R2 = 0.55, SEE = 0.008 g/cm3). The regression between % fat by HW and ADP for the group (Y = 0.84x + 3.81, R2 = 0.83, SEE = 3.35 % fat) and for the males (Y = 0.84x + 3.25, R2 = 0.90, SEE = 3.00 % fat) significantly deviated from the line of identity. However, in females the regression between % fat by HW and ADP did not significantly deviate from the line of identity (Y = 0.81x + 5.17, R2 = 0.56, SEE = 3.80 % fat). Bland-Altman analysis revealed no bias between HW total body density and ADP total body density for the entire group (R = 0.-22; P = 0.08) or for females (R = 0.02; P = 0.92), however bias existed in males (R = -0.37; P ≤ 0.05). Bland-Altman analysis revealed no bias between HW and ADP % fat for the entire group (R = 0.21; P = 0.10) or in females (R = 0.10; P = 0.57), however bias was indicated for males by a significant correlation (R = 0.36; P ≤ 0.05), with ADP underestimating % fat at lower fat values and overestimating at the higher % fat values.
Conclusion
A significant difference in total body density and % fat was observed between ADP and HW in children 10–15 years old with a potential gender difference being detected. Upon further investigation it was revealed that the study was inadequately powered, thus we recommend that larger studies that are appropriately powered be conducted to better understand this potential gender difference.
==== Body
Background
Recent results obtained from the National Health and Nutrition Examination Survey (NHANES) found increases in overweight and obesity not only in adults but also in children [1]. An alarming 31% of children aged 6 to 19 years old were found to be at risk for overweight while 16% were classified as overweight. At this time, there is no indication of this trend abating but in fact only growing worse. These facts stress the importance of accurate methods to determine body composition in a pediatric population to identify children at risk of becoming obese or those who are already obese.
Several methods can be used to determine body composition in a pediatric population, however these techniques can be costly, time consuming and difficult to administer in a pediatric population. Hydrostatic weighing (HW) has commonly been used in adults, though the administration in certain populations such as the elderly, ill, children, and certain ethnic groups has proven challenging. For successful completion of HW, multiple trials of complete head submersion followed by a maximal exhalation to record underwater weight is the likely culprit for this difficulty. In our study alone 9 children were unable to achieve the requirements necessary to complete a HW test.
Air displacement plethysmography (ADP) is a method that offers promise to alleviate problems associated with HW and is widely used to assess body composition of adults in many different settings. An ADP test involves the child sitting inside of the testing chamber while wearing a swimsuit, swim cap and nose clips for assessment of body volume with each body volume measurement lasting only 50 seconds [2]. After two successful body volume measurements, thoracic lung volume is measured using a tube placed inside of the testing chamber and using the gentle puffing maneuver. Total time to assess body composition by ADP in children is approximately 8–10 minutes compared to upwards of 30 minutes or more for HW.
Several studies have investigated the validity and feasibility of ADP in an adult population and found ADP to be valid and reliable [3-9], while others have not found agreement [10]. Although numerous studies have been completed in an adult population, a limited number of studies exist in evaluating ADP in a pediatric population [6,9,11-15] with two studies measuring the residual lung volume simultaneously [13,14].
Therefore, the purpose of this study was to evaluate body density obtained by ADP to body density obtained by HW in children and adolescents and to examine if gender differences exist between ADP and HW estimates in body density and % fat.
Methods
Subjects
A total of 77 subjects were recruited through the Norman Youth Soccer Association. Norman, Oklahoma. Two subjects' data were invalid due to instrument malfunction while an additional 9 subjects were unable to perform the HW procedures. Therefore, data from only 66 subjects were used in the analysis, of which 40 were males and 26 females between the ages of 10 – 15 years old. Subjects were excluded from participation in the study if they were claustrophobic or had any known lung disease or disorder, including asthma.
Protocol
Subjects were required to visit the Human Body Composition Laboratory at the University of Oklahoma Norman Campus for one visit with the visit lasting approximately 1 1/2 hours. ADP always preceded HW, this was done to ameliorate any residual effect that heat and moisture may have on ADP measurements [16]. Written informed consent and assent were obtained from all subjects and their parents. This research study was approved by the University of Oklahoma Institutional Review Board.
BOD POD instrumentation
The BOD POD® Body Composition System (Life Measurement Instruments, Concord, CA) was used to assess body volume and total body density with the operating procedures previously described elsewhere [2]. Total body density was converted to percent fat using the Lohman age specific equations [17]. Subjects were instructed not to eat, drink or exercise 6 hours prior to testing. All subjects wore a Speedo swimsuit (provided by the laboratory), swim cap, nose clips, and removed all jewelry prior to testing. The traditional and well established pulmonary plethysmographic measurement (used inbasic pulmonary function testing) of thoracic gas volume (TGV) is adopted by the BOD POD and incorporated into the testingfor TGV measurement and has been shown to be valid [2,18]. The only difference is that the traditional pulmonary plethysmography determines TGV at the end-tidal exhalation; however, the BOD PDO measures TGV at mid-tidal exhalation. This is done because it is necessary to correct raw body volume for the average amount of airin thelungs during normal tidal breathing, which is reflected by taking the measurement at mid-tidal exhalation. The testing procedure involved the following steps. First, the BOD POD was calibrated by computing the ratio of the pressure amplitudes (reference chamber and testing chamber) for the empty testing chamber, which is ~ 450L, and the testing chamber with our calibration cylinder (49.556-L). Following the calibration and after the TGV procedure was explained and all pertinent subject data entered into the computer software, the subject entered the testing chamber to have there raw body volume measured. After body volume was measured, the TGV was measured. The TGV was measured by having the subject sit quietly in the BOD POD while breathing through a disposable tube and filter connected to the reference chamber in the rear of the BOD POD. After four or five normal breaths and at the point of mid-exhalation the airway was occluded and the subject was instructed to make two quick and light puffs. All TGV measurements were measured and not estimated.
Hydrostatic weighing
Each child's total body density was measured by underwater weighing, with a simultaneous measurement of residual lung volume by using the closed-circuit oxygen dilution technique measurement system (EXERTECH, Dresbach, MN). The simultaneous residual lung volume measurement system included a calibrated piston pump which was used to dispense a measured volume of oxygen into a rubber bag and a fast responding electronic nitrogen gas analyzer. This continuously sampled the inhaled and exhaled gas at the subject's mouth in order to follow the nitrogen fraction of the respiratory air as it mixed with the pre-measured oxygen bag volume during re-breathing (i.e. when the subjects emerged from being submerged under the water). The nitrogen gas fraction of the mixture was continuously recorded during the re-breathing procedure and reached a relative equilibrium, usually within 5 or 6 breaths. Residual volume was calculated from the initial oxygen volume in the bag and the change in nitrogen fraction by dilution. The underwater weight was measured to the nearest 1/100th of a gram in an enclosed tile tank in which the subject, while wearing a one-piece swimsuit, sat in a carriage wrack suspended from four LCL 10 load cells integrated with a summing box and digital display calibrated from 0 to 18,000 g (Omega, Stanford, CT). After two practice trials, underwater weight and the residual lung volume were measured simultaneously 5 times. The average of multiple trial densities within 1/1000th g/cm3 were used for the underwater weight. Percent fat mass was calculated from whole body density (g/cm3) using Lohman age specific equations [17].
Data analysis
Accuracy, precision, and bias were examined in ADP with HW serving as the criterion method. Statistical significance was set at (P ≤ 0.05).
Regression analysis was used to determine the accuracy of ADP. ADP was considered to be accurate if the regression between body density and percent fat by HW and ADP did not have a slope significantly different from one and an intercept significantly different from zero. This analysis tested the hypothesis that the regression of body density and percent fat by HW and body density and percent fat by ADP did not significantly deviate from the line of identity.
The amount of shared variance between ADP and HW was assessed by R2. Potential bias between ADP and HW were examined using Bland-Altman analysis [19]. This particular test examined the difference in body density and percent fat between ADP and HW as a function of the average body density and percent fat by ADP and HW. A non-significant correlation indicated no bias was seen in the technique (i.e. ADP) across the range of fatness.
Results
The purpose of this study was to compare ADP with HW in male and female children and adolescents between the ages of 10–15 years old. Results for this study will be presented in the following order; body density findings followed by % fat findings. The physical characteristics of the subjects are presented in Table 1 and a summary of % fat means for the total group and for each gender are found in Figure 1.
Figure 1 The bars represent % fat measurements by ADP and HW for the entire group and for males and females.
Table 1 Physical Characteristics of subjects
Group (N = 66) Males (N = 40) Females (N = 26)
Age, (yr) 12.2 ± 1.6 12.4 ± 1.3 12.0 ± 1.9
Body weight, (kg) 45.0 ± 12.8 47.4 ± 14.8 41.4 ± 7.7
Height, (in) 60.7 ± 4.3 61.2 ± 4.7 59.9 ± 3.5
BMI (kg/m2) 18.7 ± 3.4 19.3 ± 4.1 17.7 ± 1.7
ADP body density (g/cm3) 1.048 ± 0.019 1.050 ± 0.023 1.045 ± 0.011
HW body density (g/cm3) 1.046 ± 0.017 1.049 ± 0.020 1.042 ± 0.011
ADP % fat 18.6 ± 8.8 18.4 ± 10.6 18.9 ± 5.2
HW % fat 19.4 ± 8.1 18.7 ± 9.4 20.4 ± 5.6
Total body density
Accuracy and the amount of shared variance between the techniques of body density was examined by regression of HW body density versus body density by ADP for the total group and for each gender. The regressions between the two methods are shown in the top panel of Figures 2, 3, and 4 for the group, males, and females respectively, while a summary of these regressions are presented in Table 2. The group (Y = 0.835x + 0.171, R2 = 0.84, SEE = 0.007 g/cm3) and male regressions (Y = 0.837x + 0.174, R2 = 0.90, SEE = 0.006 g/cm3) of ADP total body density compared to HW total body density significantly deviated from the line of identity. However, for the females the regression between ADP total body density and HW total body density did not significantly deviate from the line of identity (Y = 0750x + 0.258, R2 = 0.55, SEE = 0.008 g/cm3).
Figure 2 Top: regression of body density (g/cm3) by ADP against body density by HW for the total group. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between body density by ADP – body density by HW. The upper and lower dashed lines represents ± 2 SD from the mean. No bias between the techniques was observed as indicated by a nonsignificant P value.
Figure 3 Top: regression of body density (g/cm3) by ADP against body density by HW for males. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between body density by ADP – body density by HW. The upper and lower dashed lines represents ± 2 SD from the mean. Bias between the techniques was observed as indicated by a significant P value.
Figure 4 Top: regression of body density (g/cm3) by ADP against body density by HW for females. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between body density by ADP – body density by HW. The upper and lower dashed lines represents ± 2 SD from the mean. No bias between the techniques was observed as indicated by a nonsignificant P value.
Table 2 Summary of regression for density
R2 Intercept, (g/cm3) Slope SEE (g/cm3)
ADP group 0.84 0.171 ± 0.049* 0.835 ± 0.046** 0.007
ADP males 0.90 0.171 ± 0.048* 0.837 ± 0.045** 0.007
ADP females 0.55 0.258 ± 0.146 0.750 ± 0.140 0.008
*Significantly different from 0 (P < 0.05)
**Significantly different from 1 (P < 0.05)
A Bland-Altman analysis was performed to determine whether bias existed between ADP and HW total body density across the range of fatness for the entire group and for each gender. These analyses are presented in bottom panel of Figures 2, 3, and 4 for the entire group, males, and females respectively. For the group (R = -0.22; P = 0.08) and for females (R = 0.02; P = 0.92), no bias was observed as indicated by a non significant correlation. However, in males bias existed across the range of fatness (R = -0.37; P ≤ 0.05), with ADP overestimating total body density at lower densities and underestimating total body density at higher densities.
% Fat
A summary of % fat estimates for the entire group and for each gender for both HW and ADP are shown in Figure 1 and the summary of the regressions for % fat for the group and both genders are shown in Table 3. Accuracy of % fat was examined by the regression of HW % fat against ADP % fat for the total group and for each gender. These regressions are shown in the top panel Figures 5, 6, and 7 for the group, males, and females respectively. The regression between % fat by HW and % fat by ADP significantly deviated from the line of identity for the entire group (Y = 0.84x + 3.81, R2 = 0.83, SEE = 3.35 % fat) and in the males (Y = 0.84x + 3.25, R2 = 0.90, SEE = 3.00 % fat). However, in females the regression did not significantly deviated from the line of identity (Y = 0.81x + 5.17, R2 = 0.56, SEE = 3.80 % fat).
Table 3 Summary of regression for % fat
R2 Intercept, (kg) Slope SEE (% fat)
ADP group 0.83 3.81 ± 0.97* 0.84 ± 0.05** 3.4
ADP males 0.90 3.25 ± 0.96* 0.84 ± 0.05** 3.0
ADP females 0.56 5.17 ± 2.88 0.81 ± 0.15 3.8
*Significantly different from 0 (P < 0.05)
**Significantly different from 1 (P < 0.05)
Figure 5 Top: regression of % fat by ADP against % fat by HW for the total group. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between % fat by ADP – % fat by HW. The upper and lower dashed lines represents ± 2 SD from the mean. No bias between the techniques was observed as indicated by a nonsignificant P value.
Figure 6 Top: regression of % fat by ADP against % fat by HW for males. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between % fat by ADP – % fat by HW. The upper and lower dashed lines represents ± 2 SD from the mean. Bias between the techniques was observed as indicated by a significant P value.
Figure 7 Top: regression of % fat by ADP against % fat by HW for females. Bottom: Bland-Altman analysis for the group where the middle dashed line represents the mean difference between % fat by ADP – % fat by HW. The upper and lower dashed lines represents ± 2 SD from the mean. No bias between the techniques was observed as indicated by a nonsignificant P value.
A Bland-Altman analysis was performed to determine whether bias existed between ADP and HW across the range of body fatness for the group and for each gender. These analyses are shown in bottom panel of Figures 5, 6, and 7 for the group, males, and females respectively. ADP for the group (R = 0.21; P = 0.10) and for females (R = 0.10; P = 0.57) did not show a significant bias across the range of body fatness. ADP for males showed a significant bias across the range of body fatness (R = 0.36; P ≤ 0.05), with ADP underestimating body fat at lower fat values and overestimating at the higher fat values.
Lung volumes
A paired t – test was conducted to determine if the measured lung volume (e.g. residual lung volume in hydrostatic weighing and the thoracic gas volume in ADP) significantly deviated from its respective estimated lung volumes. A summary of the lung volumes are presented in Table 4. A significant over-estimation in the residual lung volume was observed between the measured and predicted residual lung volume for the group (1.27 ± 0.44 vs. 0.84 ± 0.12) and for both males (1.36 ± 0.48 vs. 0.86 ± 0.14) and females (1.12 ± 0.33 vs. 0.82 ± 0.10), as indicated by a P ≤ 0.05. A significant under-estimation in the thoracic gas volume was also observed between the measured and predicted thoracic gas volume, volume for the group (2.45 ± 0.68 vs. 2.71 ± 0.51) for the males, (2.53 ± 0.72 vs. 2.74 ± 0.58) and for the females (2.33 ± 0.62 vs. 2.66 ± 0.36) indicated by P ≤ 0.05.
Table 4 Summary of residual volume and thoracic gas volumes for the total group and for both genders
Residual Volume (predicted1) Residual Volume (measured) TGV (predicted2) TGV (measured)
Group 0.84 ± 0.12 1.27 ± 0.44** 2.71 ± 0.51 2.45 ± 0.68**
Males 0.86 ± 0.14 1.36 ± 0.48* 2.74 ± 0.58 2.53 ± 0.72*
Females 0.82 ± 0.10 1.12 ± 0.33* 2.66 ± 0.36 2.33 ± 0.62*
* Statistically significant from predicted (P < 0.05)
** Statistically significant from predicted (P < 0.01)
1 Polgar, 1971 RV = (0.029 × Ht(in)) - 0.919
2 Crapo, 1982 VTG = FRC + 1/2 TV
Discussion
The purpose of this study was to validate ADP with HW in a group of children of varying degrees of fatness. This study is significant because the ability to accurately measure body composition in children is challenging and difficult. With the sudden increase in the incidence of pediatric obesity, the ability to accurately determine body composition is paramount in the treatment of this escalating problem. ADP has shown to be a valid tool in an adult population [3-9,20] however, few studies have compared ADP and HW in children [6,11-15]. Thus, the purpose of this study was to compare total body density and % fat measurements obtained by ADP against those obtained by HW in male and female children between the ages of 10–15 years old.
Prior studies validating ADP and HW in children have produced varying results. Lockner et al. studied 54 children and found agreement between the two techniques by regression analysis [11]. Although regression indicated agreement between techniques, a significant group mean difference was found (P<0.0005). Dewit et al. also found no difference between total body density by HW and total body density by ADP in children ranging in age from 8–12 years old [14]. Further research comparing in children has found differences between ADP and HW. Nunez et al. studied a large sample of children consisting of 54 males and 66 females [6]. Regression analysis revealed poor agreement between total body density by HW and total body density by ADP for the entire group of children. Further analysis by Nunez et al. separated % fat findings by gender and found a significant gender difference. Percent fat measured by ADP was significantly greater than % fat measured by HW in both females and males by 1.7% and 0.5%, respectively (P < 0.0001).
Possible reasons for disagreement in the literature found between ADP and HW in children are numerous. When assessing body composition by ADP some studies predicted TGV instead of actually measuring TGV [6,11]. Research has shown using the prediction equations originally developed for adults in children will overestimate TGV resulting in inaccurate % fat measures [21]. Traditionally, measuring TGV has proven challenging in a pediatric population with 35% of pediatric centers unable to measure the TGV. It should be noted, all of the TGV measurements in this study were measured, and none were predicted. Some studies did not follow strict protocol and allowed subjects to use clothing other than a tight fitting swimsuit such as spandex bicycle shorts [6,11,22]. When a strict clothing protocol is not followed, it has been shown that % fat can be underestimated upwards of 6% [23,24]. When converting body density to % fat, the correct child specific equation must be used. Both Demerath et al. and Lockner et al. used the Siri equation to convert body density to % fat instead of the child gender specific Lohman equations [11,12]. And lastly, in a study by Lockner et al. the testing order was randomized, thus some of the children performed an ADP after performing a HW [11]. It has been shown by Fields et al., that HW prior to ADP resulted in an underestimation of % fat by approximately 3% [16].
One of the purposes of this study was to examine potential gender differences between ADP and HW. The results from this study did show significant gender difference between techniques, with males significantly deviating from the line of identity. However, this was not found in females for either total density or % fat. Nonetheless, this gender difference reported in this study should be viewed with caution because a power analysis revealed that 40 subjects per gender (Group N = 80) were needed for 80% power (alpha = 0.05) for a main effect of 1.5% fat. In this study only 26 females were used in the data analysis, though 37 were brought in for testing. This was due to the fact that 9 of the female subjects were unable to perform the HW procedures and two were lost when equipment malfunctioned. A power analysis for this data set was performed and it was found after factoring in those dropped from the study, there remained only a power of 0.429. Therefore, we did not have enough power to detect the potential difference between the methods in females.
Research has shown for ADP that prediction equations developed in adults provide invalid lung volume predictions in children. The adult prediction equations used to predict TGV were developed using a healthy adult population [25]. Fields et al. measured TGV in 113 boys and 111 girls and found prediction equations significantly overestimated TGV in both genders (P < 0.001) [21]. In the current study, residual lung volume was measured simultaneously in water and TGV was measured during testing. This is significant because only two other studies have been able to simultaneously measured residual lung volume while underwater weighing in children [13,14].
Interestingly, the coefficient of variance (CV) for repeated measures over two days in a subset of the children in this study for ADP and HW was 3.1% and 7.1% respectively. This is quite high considering the CV for ADP and HW in our laboratory for adults are 1% and 1.5% respectively, though Nunez et al. reported an ADP CV of 8.5% in a pediatric population [6]. Consequently, the high CV may be playing a role in the true relationship between ADP and HW.
Conclusion
Due to the ease in the testing procedure and high subject compliance for all ages, ADP has quickly begun to emerge as a popular body composition method to use in children and adults. In the current study, we found an overall poor agreement between ADP and HW with the study inadequately powered to make any definitive statements concerning potential gender differences between the techniques. In conclusion, we recommend more studies validating ADP and HW in children be performed utilizing a larger sample size.
Competing interests
David A. Fields has received funding from Life Measurement Incorporated for past studies (though Life Measurement Incorporated did not fund this study). None of the authors have non-financial competing interests or own stock or are applying for patents that may represent a conflict of interest.
Authors' contributions
GC carried out the hydrostatic weighing and ADP studies, participated in the design of the study, coordinated the study, and helped draft the manuscript. HH provided critical evaluation on ADP testing and participated in writing of the manuscript. DF conceived the study, performed the statistical analysis, and drafted 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 would like to acknowledge Drs. Mike Bemben and Andrew Gardner for their involvement in the project.
==== Refs
Hedley AA Ogden CL Johnson CL Carroll MD Curtin LR Flegal KM Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002 Jama 2004 291 2847 2850 15199035 10.1001/jama.291.23.2847
Dempster P Aitkens S A new air displacement method for the determination of human body composition Med Sci Sports Exerc 1995 27 1692 1697 8614327
Biaggi RR Vollman MW Nies MA Brener CE Flakoll PJ Levenhagen DK Sun M Karabulut Z Chen KY Comparison of air-displacement plethysmography with hydrostatic weighing and bioelectrical impedance analysis for the assessment of body composition in healthy adults Am J Clin Nutr 1999 69 898 903 10232628
Fields DA Wilson GD Gladden LB Hunter GR Pascoe DD Goran MI Comparison of the BOD POD with the four-compartment model in adult females Med Sci Sports Exerc 2001 33 1605 1610 11528352 10.1097/00005768-200109000-00026
Levenhagen DK Borel MJ Welch DC Piasecki JH Piasecki DP Chen KY Flakoll PJ A comparison of air displacement plethysmography with three other techniques to determine body fat in healthy adults J Parenteral Enteral Nutr 1999 23 293 299
Nunez C Kovera AJ Pietrobelli A Heshka S Horlick M Kehayias JJ Wang Z Heymsfield SB Body composition in children and adults by air displacement plethysmography Eur J Clin Nutr 1999 53 382 387 10369494 10.1038/sj.ejcn.1600735
Vescovi JD Hildebrandt L Miller W Hammer R Spiller A Evaluation of the BOD POD for estimating percent fat in female college athletes J Strength Cond Res 2002 16 599 605 12423192 10.1519/1533-4287(2002)016<0599:EOTBPF>2.0.CO;2
Vescovi JD Zimmerman SL Miller WC Hildebrandt L Hammer RL Fernhall B Evaluation of the BOD POD for estimating percentage body fat in a heterogeneous group of adult humans Eur J Appl Physiol 2001 85 326 332 11560087 10.1007/s004210100459
McCrory MA Gomez TD Bernauer EM Mole PA Evaluation of a new air displacement plethysmography for measuring human body composition Med Sci Sports Exerc 1995 27 1686 1691 8614326
Collins MA Millard-Stafford ML Sparling PB Snow TK Rosskopf LB Webb SA Omer J Evaluation of the BOD POD for assessing body fat in collegiate football players Med Sci Sports Exerc 1999 31 1350 1356 10487380 10.1097/00005768-199909000-00019
Lockner DW Heyward VH Baumgartner RN Jenkins KA Comparison of air-displacement plethysmography, hydrodensitometry, and dual X-ray absorptiometry for assessing body composition of children 10 to 18 years of age Ann N Y Acad Sci 2000 904 72 78 10865712
Demerath EW Guo SS Chumlea WC Towne B Roche AF Siervogel RM Comparison of percent body fat estimates using air displacement plethysmography and hydrodensitometry in adults and children Int J Obes Relat Metab Disord 2002 26 389 397 11896495 10.1038/sj.ijo.0801898
Fields DA Goran MI Body composition techniques and the four-compartment model in children J Appl Physiol 2000 89 613 620 10926645
Dewit O Fuller NJ Fewtrell MS Elia M Wells JC Whole body air displacement plethysmography compared with hydrodensitometry for body composition analysis Arch Dis Child 2000 82 159 164 10648375 10.1136/adc.82.2.159
Wells JC Douros I Fuller NJ Elia M Dekker L Assessment of body volume using three-dimensional photonic scanning Ann N Y Acad Sci 2000 904 247 254 10865749
Fields DA Hunter GR Higgins PB Assessment of body composition by air displacement plethysmography: Influence of body temperature and moisture Dynamic Medicine 2004 3
Lohman TG Assessment of body composition in children Pediatr Exerc Sci 1989 1 19 30
Ruppell G Manual of pulmonary function testing 1994 St. Louis, Mosby 11 14
Bland JM Altman DG Statistical methods for assessing agreement between two methods of clinical measurement Lancet 1986 1 908 909 2870372 10.1016/S0140-6736(86)91008-1
Fields DA Goran MI McCrory MA Body-composition assessment via air-displacement plethysmography in adults and children: a review Am J Clin Nutr 2002 75 453 467 11864850
Fields DA Hull HR Cheline AJ Yao M Higgins PB Child-specific thoracic gas volume prediction equations for air-displacement plethysmography Obes Res 2004 12 1797 1804 15601975
Hull HR Fields DA Effect of Short Schemes on Body Composition Measurements using Air-Displacement Plethysmography Dyn Med 2005 4 8 16045792 10.1186/1476-5918-4-8
Vescovi JD Zimmerman SL Miller WC Fernhall B Effects of clothing on accuracy and reliability of air displacement plethysmography Med Sci Sports Exerc 2002 34 282 285 11828238
Fields DA Hunter GR Goran MI Validation of the BOD POD with hydrostatic weighing: influence of body clothing Int J Obes Relat Metab Disord 2000 24 200 205 10702771 10.1038/sj.ijo.0801113
Crapo RO Morris AH Clayton PD Nixon CR Lung volumes in healthy nonsmoking adults Int Europ Physiopath Resp 1982 18 419 425
|
16153297
|
PMC1236934
|
CC BY
|
2021-01-04 16:31:06
|
no
|
BMC Pediatr. 2005 Sep 9; 5:37
|
utf-8
|
BMC Pediatr
| 2,005 |
10.1186/1471-2431-5-37
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-911613732010.1186/1471-2458-5-91Research ArticleThe impact of partial smokefree legislation on health inequalities: Evidence from a survey of 1150 pubs in North West England Tocque Karen [email protected] Richard [email protected] Brenda [email protected] North West Public Health Observatory, Centre for Public Health, Liverpool John Moores University, North Street, Liverpool, L3 2AY, UK2 Evidence for Population Health Unit, Division of Epidemiology and Health Sciences, Stopford Building, University of Manchester, Oxford Road, Manchester, M13 9PT, UK3 South Sefton Primary Care Trust, 3 rd Floor Burlington House, Crosby Road North, Waterloo, Liverpool L22 0QB, UK2005 1 9 2005 5 91 91 1 8 2005 1 9 2005 Copyright © 2005 Tocque et al; licensee BioMed Central Ltd.2005Tocque 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 UK government claims that between 10 and 30% of pubs and bars will be exempt from proposed legislation to achieve smokefree enclosed public places across England. This arises from the contentious inclusion that pubs and bars that do not prepare and serve food and private members clubs, will be able to allow smoking. We aimed to survey pubs and bars across the North West of England to assess smoking policies and the proportion and variations by deprivation level of venues preparing and serving food.
Methods
We carried out a telephone survey of 1150 pubs and bars in 14 local authorities across the North West of England. The main data items were current smoking policy, food preparation and serving status, and intention to change food serving and smoking status in the event of implementation of the proposed English partial smokefree legislation.
Results
29 pubs and bars (2.5%) were totally smoke-free, 500 (44%) had partial smoking restrictions, and 615 (54%) allowed smoking throughout. Venues situated in the most deprived quintiles (4 and 5) of deprivation were more likely to allow unrestricted smoking (62% vs 33% for venues in quintiles 1 and 2). The proportion of pubs and bars not preparing and serving food on the premises was 44% (95% CI 42 to 46%), and ranged from 21% in pubs and bars in deprivation quintile 1 to 63% in quintile 5.
Conclusion
The proportion of pubs and bars which do not serve food was far higher than the 10–30% suggested by the UK government. The proportion of pubs allowing unrestricted smoking and of non-food venues was higher in more disadvantaged areas, suggesting that the proposed UK government policy of exempting pubs in England which do not serve food from smokefree legislation will exacerbate inequalities in smoking and health.
==== Body
Background
The English Public Health White Paper 'Choosing Health' and the accompanying delivery plan [1,2] state that the government will regulate, with legislation where necessary, to achieve smokefree enclosed public places and workplaces by 2007–2008. However, pubs and bars, which do not prepare and serve food and private members clubs, will be able to allow smoking. The UK government claim exempted pubs and bars represent between 10 and 30% of the total [1].
The Choosing Health proposals raise several concerns. One is that the non-smoking exclusion zone around the bar will be ineffective and harm to the health of staff working in pubs and bars and members clubs will continue. A second is that the effect of smokefree ordinances at reducing smoking prevalence [3] will be undermined by allowing smoking to continue in key social settings like pubs, clubs and bars. Finally, pubs and bars in disadvantaged areas may be less likely to serve food and, because of a higher smoking prevalence among customers, be most likely to stop serving food to allow continued smoking. This may undermine efforts to reduce smoking prevalence in these communities, and perpetuate and exacerbate the already gross inequalities in smoking [4,5] and smoking-related ill health.
Preliminary data from a survey of 29 Local Authorities conducted by the British Medical Association suggest that the 10–30% figure for exempted premises is an underestimate, particularly outside of the south of England. Thirteen out of 29 (10 located in the north or Midlands) councils estimated that more than 30% of their pubs did not serve food [6]. Figures from monitoring of pubs by Environmental Health Officers in Northamptonshire found that 54% of pubs and bars would be exempt, 85% in Corby, the most deprived Local Authority [7]. Similarly, a recent study of the catering status of 174 pubs on the Local Authority records of Telford and Wrekin borough found that 43% of pubs would be exempt, 69% in the most deprived areas [8].
We conducted a survey across 14 Local Authorities in the North West region to assess current smoking policies, the proportion of pubs and bars preparing and serving food, and variations in this proportion by deprivation level of the local area. We hypothesised that pubs and bars from disadvantaged areas would be more likely to allow unrestricted smoking and less likely to prepare and serve food.
Methods
A structured questionnaire (available online with methodological instructions [9]) was adapted from a pilot survey carried out in Wirral Local Authority and disseminated to all 43 Local Authorities within the North West region. Fourteen Local Authorities agreed to participate: six were predominantly urban, three were predominantly rural and five were mixed urban/rural. Within each Local Authority, a local coordinator identified lists of licensed premises. Either all or a random sample of licensed premises were chosen for inclusion. At least three attempts were made to make contact with the proprietor or owner of chosen establishments. A telephone interview was conducted using the questionnaire except in a few cases where the questionnaire was completed in a face-to-face interview.
Some areas focused only on pubs, bars and clubs, whereas others also included hotel and restaurants with licensed bars. The current analysis is largely restricted to pubs and bars and private members clubs, and excludes night clubs, restaurants and hotels. Venues were allocated an area-based deprivation score (Index of Multiple Deprivation (IMD) 2004 [10]) by mapping their postcode to Census Super Output Area which were categorised into deprivation quintiles for all Super Output Areas across the north west.
Results
There were 1818 licensed pubs, bars included in the sampling frame, of which 1150 pubs and bars (63%) agreed to take part, 386 (21%) refused, and 282 (16%) could not be contacted. Of 396 members clubs contacted 195 (49%) responded and 78 (20%) refused and 123 (31%) could not be contacted. The interviewee was the manager or proprietor for 73% of pubs and bars participating.
Across the 14 Local Authorities, only 29 pubs and bars (2.5%) were totally smoke-free, 500 (44%) had partial smoking restrictions, and 615 (54%) allowed smoking throughout. Seventy-one per cent allowed smoking at the bar.
The proportion of pubs and bars not preparing and serving food on the premises was 44% (95% Confidence Interval (CI) 42 to 46%), and ranged from 19 to 55% across the 13 Local Authorities (one Local Authority excluded because only 14 venues participated). Respondents were asked if they were likely to change their policy on preparing and serving food in response to the English Public Health White Paper proposals. Of pubs and bars that currently served food, 82 (13%) stated that they were likely to stop serving food, and of those that didn't serve food 42 (9%) stated that they were likely to start serving food in response to the proposed English smokefree legislation. The net projected change in the number of venues that would prepare and serve food after introduction of the White Paper proposals was a reduction of 40 (3%), increasing the proportion of non-food pubs from 44 to 47%.
Results stratified by IMD-2004 quintiles are shown in table 1. The proportion of venues with unrestricted smoking increased with deprivation (Figure 1). In the more affluent areas only 21% (IMD-2004 quintile 1) and 40% (quintile 2) allowed smoking throughout compared with 56% in deprivation quintile 4 and 67% in the most deprived quintile 5 – the absolute difference in proportions between venues in quintiles 1–2 and quintile 4–5 was 28%, (95% CI 21 to 36%).
Table 1 Smoking and food preparation policies by IMD-2004 quintiles of deprivation in pubs and bars in 14 Local Authorities in the North West of England.
All (n = 1150) IMD-2004 Quintile IMD-2004 Missing (n = 70)
1 (n = 70) 2 (n = 126) 3 (n = 180) 4 (n = 355) 5 (n = 349)
% with unrestricted smoking 53.8 21.4 39.7 47.2 56.3 67.1 49.3
% allowing smoking at the bar 70.7 55.7 66.7 68.3 71.8 77.9 57.1
% currently not preparing and serving food 43.6 21.4 28.6 38.9 36.9 63.3 40.0
% not preparing and serving food post White Paper (predicted) 47.0 20.0 31.8 41.1 46.2 62.8 41.4
% of food pubs allowing smoking in food serving areas 75.7 (n = 649) 80.0 (n = 55) 74.4 (n = 90) 76.4 (n = 110) 75.0 (n = 224) 67.2 (n = 128) 64.3 (n = 42)
% of pubs allowing children on premises 60.3 77.1 79.4 69.4 52.1 54.2 57.1
% of pubs allowing children on premises with smoking in children's areas 81.4 (n = 693) 85.2 (n = 54) 83.0 (n = 100) 81.6 (n = 125) 75.7 (n = 185) 78.3 (n = 189) 75.0 (n = 40)
The Index of Multiple Deprivation 2004 (IMD 2004) is a Super Output Area (SOA) level measure of multiple deprivation and is made up of seven SOA level Domain Indices: Income deprivation, Employment deprivation, Health deprivation and disability, Education, skills and training deprivation, Barriers to Housing and Services, Living environment deprivation and Crime. The overall IMD is conceptualised as a weighted area level aggregation of these specific dimensions of deprivation [12].
Figure 1 Current smoking policies in North West pubs and bars by deprivation quintile (IMD 2004). Data showing percent of venues with 95% confidence intervals. IMD quintile was allocated by the Census Super Output Area (SOA) of the venue as determined by the postcode of the venue. Quintile 1 represents the most affluent 20% of SOAs and quintile 5 represents the most deprived 20% of SOAs in the North West region.
The proportion of pubs and bars not preparing and serving food increased with the level of deprivation (Figure 2) from: 21% in the most affluent areas (IMD-2004 quintile 1) to 37–39% in quintiles 3 and 4, and 63% in quintile 5 (χ2 = 65.1, p < 0.001). Difference in proportions between quintiles 1 and 5 was 42% (95% CI 30% to 51%). Based on respondents' stated intentions, the English Public Health White Paper proposals are likely to exaggerate further the gradient in the proportion of non-food pubs and bars by deprivation level (Figure 2).
Figure 2 Current and predicted preparation and serving of food in North West pubs and bars by deprivation quintile (IMD 2004). Data showing percent of venues with 95% confidence intervals. IMD quintile was allocated by the Census Super Output Area (SOA) of the venue as determined by the postcode of the venue. Quintile 1 represents the most affluent 20% of SOAs and quintile 5 represents the most deprived 20% of SOAs in the North West region.
Over half of the venues allowed children on the premises. This proportion was lowest in venues in the most disadvantaged areas. Around 80% of pubs and bars where children were allowed permitted smoking in the children's areas. This varied little by IMD-2004 quintile (Table 1). Over three quarters of pubs which prepared and served food allowed smoking in the areas where food was served. This was slightly less in venues in the most deprived quintiles.
Of licensed members clubs, 6% were smokefree, 33% had partial smoking restrictions and 61% allowed unrestricted smoking. There was a gradient in the proportion allowing unrestricted smoking, from 31% in IMD-2004 quintile 1 to 68% in quintile 5.
Discussion
We found that 44% of pubs and bars across 14 Local Authorities in the North West of England do not prepare and serve food – far higher than the 10–30% claimed by the UK government [1]. Furthermore, more respondents indicated that they would stop rather than start serving food in response to the White Paper proposals, resulting in a projected increase of 3% in the proportion not preparing and serving food.
There was a strong socio-economic gradient in the distribution of non-food serving pubs and bars: with 63% of businesses located in the most deprived areas (quintiles 4 and 5) not preparing and serving food. Unrestricted smoking was also much more likely in pubs, bars and members clubs in the most deprived areas.
This is the largest and most comprehensive survey of the likely impact of the English smokefree proposals from over 1100 pubs and bars from a wide range of Local Authorities across the North West. It explores variations by level of deprivation below Local Authority level of aggregation and is the first to examine likely changes to food serving policy after the proposed legislation is implemented. The findings are based on data collected directly from pubs and bars rather than estimates made by Local Authorities [6].
The White Paper Smokefree Consultation document [11] notes that it has been suggested that the 'Choosing Health' proposals will result in smoking pubs and bars being concentrated in deprived communities, thereby exacerbating health inequalities. This is confirmed by our survey, which found that pubs and bars that don't serve food and hence will be able to allow smoking after 2008 are more concentrated in disadvantaged areas in the North West. It is highly probable that this socio-economic gradient in the location of food/non-food serving establishments will also exist in other parts of England.
Conclusion
The impact of the 'Choosing Health' proposals are likely to contribute to maintaining the huge inequalities in smoking prevalence and smoking-related morbidity and mortality by perpetuating a strong smoking culture, reducing the impact of cessation in response to smokefree policies, and maximizing exposure of bar staff and non-smoking customers to passive smoking in the most deprived areas. All these work against the stated Government objective of reducing health inequalities due to smoking.
Competing interests
The author(s) declare that they have no competing interests. Richard Edwards is unpaid chair of North West ASH and Brenda Fullard is seconded to the North West Public Health Team, Government Office for the North West.
Authors' contributions
Following a pilot survey carried out by Tina Williams at Smokefree Wirral, KT, RE and BF developed the idea for a regional survey and modified the questionnaire. KT designed the data collection system, co-ordinated the distribution of the questionnaire and collation of the data. Data collection was organised by several local co-ordinators. KT and RE planned and carried out the analysis. The paper was drafted initially by RE, and then developed further with contributions from all the co-authors.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We would like to thank Tina Williams and Smokefree Wirral for a pilot survey and providing the original questionnaire. We are also extremely grateful to all the Local Authority, Primary Care Trust and University staff who helped carry out the survey in the following areas across the North West region: Blackburn with Darwen, Burnley, Bury, Carlisle, Eden, Hyndburn, Liverpool, Manchester, Pendle, Preston, Ribble Valley, Rochdale, Rossendale, and Salford. Particular thanks go to: Richard Holford, Kim Hargreaves, Claire Probert, Nikolas Storey, Lynne Ratcliffe, Barbara Bellis, Aaron Barker and Mark A Bellis.
The study was carried out and supervised by NHS and University-funded staff. No specific funding was provided for the study from any source. The researchers were independent of the supporting organisation bodies, who had no influence over the design, conduct, analysis or dissemination of the results of the study.
==== Refs
Department of Health Choosing health: making healthier choices easier 2004 London, Stationary Office
Department of Health Delivering choosing health: making healthier choices easier 2005 London, Stationary Office
Fichtenberg CM Glantz SA Effect of smoke-free workplaces on smoking behaviour: systematic review BMJ 2002 325 188 194 12142305 10.1136/bmj.325.7357.188
Jarvis M Wardle J Marmot M and Wilkinson RG Social patterning of individual health behaviours: the case of cigarette smoking Social Determinants of Health 1999 Oxford, Oxford University Press
Richardson K Crosier A Smoking and health inequalities 2002 Action on Smoking and Health, Health Development Agency
British Medical Association Booze, fags and food 2005 BMA
Charter NSF Survey data for pubs exempt from smokefree proposals in Choosing Health 2005 Northampton, Northampton Primary Care Trust
Woodall AA Sandbach EJ Woodward CM Aveyard P Merrington G The partial smoking ban in licensed establishments and health inequalities in England: modelling study BMJ 2005 16113033
Tobacco Control Research Programme Centre for Public Health 2005 Liverpool John Moores University
Office of the Deputy Prime Minister The English Indices of Deprivation 2004: Summary (revised) 2004 London, The Stationery Office
Department of Health Consultation on the smokefree elements of the Health Improvement and Protection Bill 2005 London
Indices of deprivation 2004 Office of the Deputy Prime Minister
|
16137320
|
PMC1236935
|
CC BY
|
2021-01-04 16:28:57
|
no
|
BMC Public Health. 2005 Sep 1; 5:91
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-91
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-941615688910.1186/1471-2458-5-94Research ArticleValidation of self reported diagnosis of hypertension in a cohort of university graduates in Spain Alonso Alvaro [email protected] Juan José [email protected]íguez Miguel [email protected]ínez-González Miguel Angel [email protected] Department of Preventive Medicine and Public Health, University of Navarra, Pamplona, Spain2 Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA3 Service of Internal Medicine, University Clinic, University of Navarra, Pamplona, Spain4 Department of Health Sciences, University of Jaén, Jaén, Spain2005 12 9 2005 5 94 94 2 6 2005 12 9 2005 Copyright © 2005 Alonso et al; licensee BioMed Central Ltd.2005Alonso 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 search for risk factors of hypertension requires the study of large populations. Sometimes, the only feasible way of studying these populations is to rely on self-reported data of the outcome. The objective of this study was to evaluate validity of self-reported diagnosis of hypertension in a cohort of university graduates in Spain.
Methods
The Seguimiento Universidad de Navarra (SUN) Study is a cohort of more than 15,000 university graduates in Spain. We selected a random sample of 79 cohort participants who reported a diagnosis of hypertension and 48 participants who did not report such diagnosis (76% participation proportion). Then, we compared information on the self-reported diagnosis of hypertension and hypertension status as assessed through two personal blood pressure measurements and an interview. Additionally, we compared self-reported and measured blood pressure levels with intraclass correlation coefficients and the survival-agreement plot.
Results
From those 79 reporting a diagnosis of hypertension, 65 (82.3%, 95% CI 72.8–92.8) were confirmed through conventional measurement of blood pressure and the interview. From those 48 that did not report a diagnosis of hypertension, 41 (85.4%, 95% CI 72.4–89.1) were confirmed as non hypertensives. Results were similar among men and women, but were worse for overweight and obese individuals, and for those with a family history of hypertension. The agreement between self-reported and measured blood pressure levels (as a continuous variable), as estimated by the intraclass correlation coefficient, was 0.35 for both systolic and diastolic blood pressure.
Conclusion
Self-reported hypertension among highly educated participants in a cohort study is a relatively valid tool to assess the hypertensive status of participants. However, the investigators should be cautious when using self-reported blood pressure values.
==== Body
Background
High blood pressure or hypertension (HT) is a major health problem in our environment.[1] The low degree of awareness among the population and the difficulties to comply with prescribed treatments, stress the importance of primary prevention of this disease.[2]
The search for risk factors of incident HT requires the study of large populations. Sometimes, the only feasible way of studying these populations is to rely on self-reported data of the outcome. As a consequence, it is of the utmost importance to assess the validity of that information.
The validity of the self-reported diagnosis of HT has been assessed in several populations, including a subsample of the EPIC-Spain cohort.[3] Results vary depending on the population and on the gold standard used (conventional measurement of blood pressure (BP) or examination of medical records).
Our objective was to assess the validity of self-reported diagnosis of HT in a random sample of the participants in the Seguimiento Universidad de Navarra (SUN, University of Navarra Follow-up) study, a cohort study in Spain.
Methods
The SUN Study
The SUN Study is a dynamic cohort of university graduates, recruited and followed up through mailed questionnaires. The main objective of the study was to assess the association between a Mediterranean dietary pattern and the risk of cardiovascular disease, diabetes, obesity, and HT. Its methods have been extensively described elsewhere.[4] Briefly, beginning in December 1999, all graduates of the University of Navarra, Registered Nurses in Navarra, and members of other professional associations received a questionnaire and a letter of invitation, explaining the objectives and design of the study. At December 2004, 17,500 had answered the initial questionnaire, and the recruitment is permanently open. Every other year, a follow-up questionnaire is mailed to each participant, gathering information about new medical diagnosis and changes in exposures of interest. The SUN Study was approved by the Institutional Review Board of the University of Navarra, and conforms to the principles embodied in the Declaration of Helsinki.
Questionnaires
The baseline questionnaire gathered information about sociodemographic variables, anthropometric measures (weight, height), lifestyle factors (smoking, physical activity), diet, and clinical variables. The participants were asked whether they had ever received a medical diagnosis of HT, their habitual use of medications, and their most recent BP measurement (choosing among the following categories, in mm Hg: lower than 100, 101–110, 111–120, 121–130, 131–140, 141–150, 151–160, 161–175, greater than 175 for systolic BP; lower than 60, 61 to 70, 71 to 80, 81 to 90, 91 to 100, 101 to 110, 111 to 120, 121 to 130, greater than 130 for diastolic BP). This question did not differentiate between casual BP determinations or more formal BP measurements carried out according to diagnostic protocols.
The follow-up questionnaire inquired about new diagnosis of HT asking whether the participant had been diagnosed by a physician since the last questionnaire.
Validation study
In September 2003, there were 2,929 SUN participants living in the metropolitan area of Pamplona (postal codes beginning by 310). Among them, 151 referred to be hypertensive (5.2%). Based in results from the literature, [3,5,6] we assumed that 80% of them would be true cases of HT. In order to obtain an 8% precision in the estimates and expecting 10% of non-response, we selected a random sample of 107 individuals that referred a diagnosis of HT in the baseline or in the two-year follow-up questionnaires, residing in the metropolitan area of Pamplona, that were alive and did not participate in a previous study on validation of diet and physical activity information. Similarly, assuming that 90% of those not reporting HT would be true normotensives, an 8% precision in the estimate and a non-response rate of 10%, we randomly selected 61 individuals, with the same inclusion/exclusion criteria than for the self-reported hypertensives.
We sent them a letter with the objectives of the validation study, an informed consent form and a contact information form (e-mail address, telephone number and hours to be called), together with a postage-paid envelope. After three months, non-respondents were sent a second or third mailing if needed. Finally, we tried to contact non-respondents by phone or email. A hundred and fifty two (response rate 90.5%) individuals accepted to participate in the validation study, 97 (response rate 90.7%) among the hypertensives and 55 among normotensives (response rate 90.2%).
After the participants gave their written consent, an appointment was made at their home, working place, or the Check up Unit at the University Clinic for the BP measurement. A structured interview was done including two BP measurements and a questionnaire about medication use and lifestyle issues related to HT. Two medical doctors (AA, JJB) carried out the field work, including the BP measurements, from September 2003 to November 2004. At the time of the measurement, both study physicians were unaware of the hypertensive status of the participant as defined in the questionnaire.
During the first minutes of the interview and with both the participant and the investigator sited down, the investigator explained the participant the objectives of the study, the type of interview, the BP measurements procedure and the confidentiality of the information. Then, the first BP measurement was done using an automatic BP measurement device Omrom M4-I. This device has been previously validated. [7] After another five minutes, used to complete the rest of the questionnaire, the second BP measurement was done. Hypertensive patients under drug therapy were not asked to stop using antihypertensive medication, because current use of antihypertensive medication was considered confirmatory of being true hypertensives.
A total of 127 participants (83.6%) completed the validation protocol. The remaining 25 (7 normotensives and 18 hypertensives) were lost either because they changed their contact information and could not be located, they refused to participate, failed to make an appointment with the investigator or had changed the place of residency out of the region and could not make an appointment in Pamplona. Final participation among hypertensives and normotensives was, respectively, 78.7% and 73.8%.
Definition of self-reported HT and 'true' HT
We considered a participant had self-reported HT when s/he answered to have been diagnosed as hypertensive by a physician either in the basal or in the follow-up questionnaire. Otherwise, s/he was considered as non-hypertensive.
We considered a participant as true hypertensive when the average of both BP measurements was ≥140 mmHg for systolic BP and/or ≥ 90 mmHg for diastolic BP, when s/he was currently using antihypertensive drug treatment or when s/he presented a medical report with a diagnosis of HT.[8]
Statistical analysis
We computed the proportion of confirmed cases of HT as the number of those who reported a diagnosis of HT and had HT according to our gold standard, divided by all those reporting a diagnosis of HT. Similarly, we computed the proportion of confirmed non hypertensives as the number of those who did not report a diagnosis of HT and were non hypertensives according to our gold standard, divided by the total number of individuals non reporting a HT diagnosis. We studied agreement between self-reported and measured BP using a random-effects model intraclass correlation coefficient [9] and the survival-agreement plot proposed by Luiz et al.[10] In the survival-agreement plot, the absolute difference Xi between BP measures was plotted in the x-axis against the proportion of pair of observations with an absolute difference equal or lower than Xi using the Kaplan-Meier method.[10] We also used the modification proposed by Llorca and Delgado to detect bias in any of the measurement methods.[11] According to the proposed modification, we separated those observations with self-reported BP higher than measured BP, and those with measured BP higher than self-reported BP. Then, we compared absolute differences both groups using the log-rank test.
To compute the sensitivity and the specificity of the self-reported diagnosis of HT, we estimated the expected distribution of true and false positives and negatives in the sampled population, based on the sampling fractions and the observed percentages of confirmed diagnosis. Then, we computed the kappa coefficient and the true prevalence of HT in that population. Confidence interval (CI) for the prevalence of HT was estimated as suggested by Cochran for stratified sampling.[12]
Results
We included 70 men and 57 women in our analyses. Mean age was 53 among those self-reporting HT and 37 among those not reporting a HT diagnosis (range 22–83 and 23–72 respectively). A total of 60 (47.3%) had a BMI ≥ 25 kg/m2. Table 1 shows the main characteristics of the study participants.
Table 1 Characteristics of participants in the validation study by self-reported HT status.
Self-reported HT (n = 79) Non self-reported HT (n = 48)
Women (%) 36.8 54.2
Age
≤ 40 y (%) 16.5 68.8
41–55 y (%) 30.4 22.9
>55 y (%) 53.2 8.2
Body mass index
<25 kg/m2 (%) 40.5 72.9
≥ 25 kg/m2 (%) 59.5 27.1
Family history of HT (% yes) 45.6 27.1
We confirmed 65 (82.3%) of the 79 self-reported HT cases (95% CI 72.8–92.8%). Among 48 participants who did not report a HT diagnosis in the questionnaires, 41 (85.4%, 95% CI 72.4–89.1%) could be considered normotensives according to our gold standard (table 2). In this last group, when the cut-off point for HT was 160/95 instead of 140/90, the proportion of confirmed normotensives increased to 97.9% (95% CI 88.7% to 100%).
Table 2 Hypertension status and validity of self-reported hypertension according to relevant variables
N (%) % Confirmed HT (95% CI)§ P-value* % Confirmed non-HT (95% CI)† P-value*
Total 127 (100) 82.3 (72.8–92.8) 85.4 (72.4–89.1)
Men 70 (55.1) 85.4 (72.8–92.8) 0.53 77.3 (56.6–89.9) 0.22
Women 57 (44.9) 78.6 (60.5–89.8) 92.3 (75.9–97.9)
Age 0.27 0.73
≤ 40 46 (36.2) 69.2 (42.4–87.3) 87.9 (72.7–95.2)
41–55 35 (27.6) 79.2 (59.5–90.8) 81.8 (52.3–94.9)
>55 46 (36.2) 88.1 (75.0–94.8) 75.0 (30.1–95.4)
Body mass index (kg/m2) 0.84 0.37
<25 67 (52.7) 81.3 (64.9–71.1) 88.6 (74.0–95.5)
≥ 25 60 (47.3) 83.0 (70.9–91.1) 76.9 (49.7–91.8)
Family history of HT 0.01 0.004
No 78 (61.4%) 72.1 (57.3–83.3) 94.3 (81.4–98.4)
Yes 49 (38.6%) 94.4 (81.9–98.5) 61.5 (35.5–82.3)
Biomedical degree 0.72 0.32
No 101 (61.4%) 81.0 (69.6–88.8) 81.6 (66.6–90.8)
Yes 26 (38.6%) 87.5 (64.0–96.5) 100.0 (72.2–100.0)
HT: hypertension. CI: confidence interval.
§True positives (according to both methods)/Positives (according to self-reporting)
† True negatives (according to both methods)/Negatives (according to self-reporting)
* Pearson's chi squared
There were no antihypertensive drug users among those reporting no hypertension, and 46% of those reporting a diagnosis of hypertension (36 out of 79) were taking antihypertensive drugs at the time of the interview. Among the remaining 43, only 14 (33%) had their BP measurements under 140/90 and were not receiving drug treatment for hypertension.
The proportion of confirmed hypertensives was higher among those groups with an expected higher prevalence of HT (men, older people, and those with high BMI or with a family history of HT). The proportion of confirmed normotensives followed an inverse pattern (table 2).
Taking into account our sampling fractions and assuming our estimate for the proportion of confirmed hypertensives, we expected that 124 out of the 151 individuals reporting a medical diagnosis of HT in the source population were true hypertensives (true positives) and 27 were normotensives (false positives).
Likewise, the number of true normotensives (true negatives) in the source population would be 2373 (from 2778 self-reported normotensives) and 405 would be hypertensives (false negatives). Based on these assumptions, the values for sensitivity, specificity and kappa coefficient would be 0.23, 0.99, and 0.31, respectively. The prevalence of HT in this population was 18.1%.
In spite of the categorization used to collect self-reported data about BP (see above), we calculated the intraclass correlation coefficient and its 95% CI to assess the absolute agreement between self-reported BP and directly measured BP as a continuous variable (Table 3). In general, the correlation between self-reported and directly observed information was low, similar for systolic and diastolic BP and higher for men than women.
Table 3 Intraclass correlation coefficients (95% CI) between self-reported Blood Pressure* and directly measured blood pressure §
Systolic BP Diastolic BP
Total 0.35 (0.09–0.55) 0.35 (0.16–0.51)
Men 0.36 (0.01–0.62) 0.45 (0.20–0.64)
Women 0.30 (0.01–0.54) 0.24 (-0.07–0.51)
≤ 55 years 0.29 (0.02–0.52) 0.23 (-0.04–0.47)
>55 years 0.27 (-0.03–0.53) 0.41 (0.12–0.63)
BP: blood pressure. CI: confidence interval.
* Seven categories for systolic BP (from <100 to >150 mmHg) and 7 categories for diastolic BP (<60 to >130) were offered for participants' choice in the questionnaire. We assigned to each category its middle point (95 to 155 mmHg for systolic and 55 to 135 mmHg for diastolic BP) in order to compute the correlation coefficients.
§ The average of two measurements taken 5 minutes apart.
Finally, we used the survival-agreement plot to depict graphically the agreement between self-reported and BP measurements (Figure 1). Using the modification of this method proposed by Llorca and Delgado to detect bias, we noted that measured systolic BP tended to be higher than self-reported systolic BP (log-rank test, p = 0.0005). However, this bias was not apparent for diastolic BP (log-rank test, p = 1.00).
Figure 1 Survival-agreement plot, as proposed by Luiz et al.[10] The x-axis shows the absolute difference between self-reported and measured blood pressure (BP), and the y-axis shows the proportion of observations with differences that are at least the observed difference. Separate lines for systolic and diastolic BP.
Discussion
Our findings showed an acceptable degree of confirmation of self-reported diagnoses of HT, but the overall agreement was not very high. Particularly, our assumptions for HT prevalence and taking into account the sampling fractions, the sensitivity and the kappa coefficient were low. On the other side, specificity was excellent. Although these results may seem discouraging, we should consider that the use of self-reported diagnoses with low sensitivity but very high specificity in a cohort study do not represent a substantial drawback, because it is very likely that HT, as other chronic diseases, will end eventually showing themselves up during the follow-up of participants. Then, in this particular setting, it would be more important to retain a high specificity, hoping that in the long-term new cases of HT will eventually be diagnosed. In addition, it is important not to forget that all except one of our false negative cases disappeared when the cut-off point for HT was 160/95 instead of 140/90 mmHg, and that there were no individuals taking antihypertensive medication among those reporting normal blood pressure.
Several studies with different methodology have evaluated the validity of self-reported diagnosis of HT. For example, in the EPIC-Murcia cohort, the kappa coefficient between self-reported and medical record-based diagnosis of HT was 0.58, but the investigators did not personally measured the BP of participants, as we did, because their gold standard were only the clinical records.[3] In the South Carolina Cardiovascular Disease Prevention Project, the sensitivity, specificity and positive and negative predictive values were, respectively, 79, 91, 76 and 93 for white women, and 62, 91, 75, and 85 for white men, with no differences between overweight and normal weight subjects.[13] In a sample of Finnish individuals, self-reported HT was confirmed reviewing medical records, obtaining similar results.[6] In the National Health and Nutritional Examination Survey III, the sensitivity for the self-reported diagnosis of HT was 71% and the specificity 90%.[14] Other studies have found similar results. [15-18] Finally, in the Nurses' Health Study and the Health Professional Follow-up Study, with a design similar to the SUN Study, the observed concordance rates among true HT diagnosis and self-reported cases of HT were comparable to ours. [5,19]
Our study has several drawbacks. First, the number of study subjects was relatively small and, thus, validity estimations had wide confidence intervals. Particularly, the separate analysis for different subgroups should be interpreted cautiously. Second, our 'gold standard', two isolated BP measurements, has a limited validity. Actually, HT diagnosis should be based on multiple BP measurements, taken on separate occasions. [8,20] Third, our study design did not allow the direct computation of confidence intervals for sensitivity, specificity and the kappa statistic. On the other side, the high educational level of our study participants ensures that health care utilization and, consequently, HT diagnosis are not influenced by educational status. And, finally, the physicians that performed the BP measurements were unaware of the questionnaire answers, making both assessments of HT diagnosis completely independent, a condition required for validation studies.
The observed agreement between observed and self-reported values of systolic and diastolic BP, as expressed by the intraclass correlation coefficient and the survival-agreement plot, was not high. However, BP levels have a high within-person variability and, in fact, BP is very difficult to track in a population (tracking being defined as the stability of a certain variable over time or the predictability of later values from earlier measurements).[21] In fact, systolic BP measurements tended to be higher than self-reported BP in our population, probably due to a real increase in BP levels over time and also due to a possible white-coat effect.[22]
Finally, we acknowledge that some misclassification will always exist in the self-reported diagnosis of HT. But, on the other side, the study of large populations would be unfeasible if we could only rely on conventional measurements, given the high amount of resources required to perform an accurate diagnosis of HT. The trade-off between precision and sample size has to be kept in mind.
Conclusion
In conclusion, self-reported HT diagnosis in the SUN Study participants showed enough validity as to be used in this large cohort study. However, our results do not support the use of self-reported BP levels (i.e. a continuous variable) as a valid measurement of usual BP levels.
List of abbreviations
BP: blood pressure
CI: confidence interval
HT: hypertension
SUN: Seguimiento Universidad de Navarra, University of Navarra Follow-up Study
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AA participated in the study design, the acquisition of data, the study analysis, and the first drafting of the manuscript. JJB participated in the acquisition of data and in the interpretation of results. MDR participated in the study design and in the interpretation of data, and provided statistical expertise. MAM have made substantial contributions to conception and design of the study, participated in the statistical analysis and the interpretation of data. All authors participated have revised the manuscript for important intellectual content and read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors thank the collaboration of the participants in the SUN Study. The SUN Study has received funding from the Spanish Ministry of Health (Grants PI030678, PI040233 and G03/140), the Navarra Regional Government (PI41/2005) and the University of Navarra. Dr. Alonso was supported partially by a Fulbright fellowship and a MMA Foundation grant.
==== Refs
Wolf-Maier K Cooper RS Banegas JR Giampaoli S Hense HW Joffres M Kastarinen M Poulter NR Primatesta P Rodríguez-Artalejo F Stegmayr B Thamm M Tuomilehto J Vanuzzo D Vescio F Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States Journal of the American Medical Association 2003 289 2363 2369 12746359 10.1001/jama.289.18.2363
Whelton PK He J Appel LJ Cutler JA Havas S Kotchen TA Roccella EJ Stout R Vallbona C Winston MC Karimbakas J for the National High Blood Pressure Education Program Coordinating Committee Primary prevention of hypertension: clinical and public health advisory from the National High Blood Pressure Education Program Journal of the American Medical Association 2002 288 1882 1888 12377087 10.1001/jama.288.15.1882
Tormo MJ Navarro C Chirlaque MD Barber X Validation of self diagnosis of high blood pressure in a sample of the Spanish EPIC cohort: overall agreement and predictive values. Epic Group of Spain J Epidemiol Community Health 2000 54 221 226 10746117 10.1136/jech.54.3.221
Martínez-González MA Sánchez-Villegas A de Irala-Estévez J Martí A Martínez JA Mediterranean diet and stroke: objectives and design of the SUN Project Nutritional Neuroscience 2002 5 65 73 11929200 10.1080/10284150290007047
Colditz GA Martin P Stampfer MJ Willett WC Sampson L Rosner B Hennekens CH Speizer FE Validation of questionnaire information on risk factors and disease outcomes in a prospective cohort study of women American Journal of Epidemiology 1986 123 894 900 3962971
Haapanen N Miilunpalo S Pasanen M Oja P Vuori I Agreement between questionnaire data and medical records of chronic diseases in middle-aged and elderly Finnish men and women American Journal of Epidemiology 1997 145 762 769 9126003
Artigao M Llavador J Rubio M Torres C López J Sanchís C Vidal A Puras A Evaluation of two devices for self-measurement of blood pressure according to the British Hypertension Society protocol: The Omrom M4 and MX2 American Journal of Hypertension 1998 16 S276
Chobanian AV Bakris GL Black HR Cushman WC Green LA Izzo JLJ Jones DW Materson BJ Oparil S Wright JTJ Roccella EJ National High Blood Pressure Education Program Coordinating Committee The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: The JNC 7 Report JAMA 2003 289 2560 2571 12748199 10.1001/jama.289.19.2560
Shrout PE Fleiss JL Intraclass correlations: uses in assessing rater reliability Psychological Bulletin 1979 86 420 428 10.1037//0033-2909.86.2.420
Luiz RR Leal Costa AJ Kale PL Werneck GL Assessment of agreement of a quantitative variable: a new graphical approach Journal of Clinical Epidemiology 2003 56 963 967 14568627 10.1016/S0895-4356(03)00164-1
Llorca J Delgado-Rodríguez M Survival techniques were used to assess agreement of a quantitative variable Journal of Clinical Epidemiology 2005 58 314 315 15718122 10.1016/j.jclinepi.2004.10.011
Cochran WG Cochran WG Stratified random sampling Sampling Techniques 1977 3rd New York, John Wiley and Sons 89 114
Giles WH Croft JB Keenan NL Lane MJ Wheeler FC The validity of self-reported hypertension and correlates of hypertension awareness among blacks and whites within the stroke belt American Journal of Preventive Medicine 1995 11 163 169 7662395
Vargas CM Burt VL Gillum RF Pamuk ER Validity of self-reported hypertension in the National Health and Nutrition Examination Survey III, 1988-1991 Preventive Medicine 1997 26 678 685 9327477 10.1006/pmed.1997.0190
Bush TL Miller SR Golden AL Hale WE Self-report and medical record report agreement of selected medical conditions in the elderly American Journal of Public Health 1989 79 1554 1556 2817172
Kehoe R Wu SY Leske MC Chylack LTJ Comparing self-reported and physician reported medical history American Journal of Epidemiology 1994 139 813 818 8178794
Martin LM Leff M Calonge N Garrett C Nelson DE Validation of self-reported chronic conditions and health services in a managed care population American Journal of Preventive Medicine 2000 18 215 218 10722987 10.1016/S0749-3797(99)00158-0
Bergmann MM Jacobs EJ Hoffmann K Boeing H Agreement of self-reported medical history: comparison of an in-person interview with a self-administered questionnaire European Journal of Epidemiology 2004 19 411 416 15233312 10.1023/B:EJEP.0000027350.85974.47
Ascherio A Rimm EB Giovanucci EL Colditz GA Rosner B Willett WC Sacks FM Stampfer MJ A prospective study of nutritional factors and hypertension among US men Circulation 1992 86 1475 1484 1330360
European Society of Hypertension - European Society of Cardiology Guidelines Committee 2003 European Society of Hypertension - European Society of Cardiology guidelines for the management of arterial hypertension Journal of Hypertension 2003 21 1011 1053 12777938
Wilsgaard T Jacobsen BK Schirmer H Thune I Lochen ML Njolstad I Arnesen E Tracking of cardiovascular risk factors. The Tromso Study, 1979-1995 American Journal of Epidemiology 2001 154 418 426 11532783 10.1093/aje/154.5.418
Hoegholm A Kristensen KS Madsen NH Svendsen TL White coat hypertension diagnosed by 24-h ambulatory monitoring. Examination of 159 newly diagnosed hypertensive patients American Journal of Hypertension 1992 5 64 70 1550667
|
16156889
|
PMC1236936
|
CC BY
|
2021-01-04 16:28:57
|
no
|
BMC Public Health. 2005 Sep 12; 5:94
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-94
|
oa_comm
|
==== Front
BMC Pulm MedBMC Pulmonary Medicine1471-2466BioMed Central London 1471-2466-5-101613732310.1186/1471-2466-5-10Research ArticleInfluence of condensation temperature on selected exhaled breath parameters Goldoni Matteo [email protected] Andrea [email protected] Roberta [email protected] Diana [email protected] Paola [email protected] Maria Vittoria [email protected] Massimo [email protected] Antonio [email protected] National Institute of Occupational Safety and Prevention, Research Center at the University of Parma, Parma, Italy2 Department of Clinical Medicine, Nephrology and Health Sciences, University of Parma, Parma, Italy2005 1 9 2005 5 10 10 9 5 2005 1 9 2005 Copyright © 2005 Goldoni et al; licensee BioMed Central Ltd.2005Goldoni 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 effects of changes in cooling temperature on biomarker levels in exhaled breath condensate have been little investigated. The aim of the study was to test the effect of condensation temperature on the parameters of exhaled breath condensate and the levels of selected biomarkers.
Methods
Exhaled breath condensate was collected from 24 healthy subjects at temperatures of -10, -5, 0 and +5 C degrees. Selected parameters (condensed volume and conductivity) and biomarkers (hydrogen peroxide, malondialdehyde) were measured.
Results
There was a progressive increase in hydrogen peroxide and malondialdehyde concentrations, and condensate conductivity as the cooling temperature increased; total condensate volume increased as the cooling temperature decreased.
Conclusion
The cooling temperature of exhaled breath condensate collection influenced selected biomarkers and potential normalizing factors (particularly conductivity) in different ways ex vivo. The temperature of exhaled breath condensate collection should be controlled and reported.
==== Body
Background
Exhaled breath condensate (EBC) is a biological fluid that mainly consists of water, but also contains small droplets of airway lining fluid. Much of the interest of EBC lies in the fact that its collection is totally non-invasive and does not lead to any discomfort or risk [1]. It has been used to assess inflammatory airway diseases such as asthma [2], chronic obstructive pulmonary disease [3], lung cancer [4], interstitial lung disease [5] and acute respiratory distress syndrome [6], and has recently also been extended to the biological monitoring of workers exposed to cobalt and tungsten [7].
EBC contains both volatile and non-volatile substances. Volatile or semi-volatile substances have appreciable vapour pressure at body temperature, and can therefore be breathed out as gases; furthermore, volatile substances in gaseous phase can be dissolved in condensed water during EBC collection depending on their physico-chemical properties [8]. Non-volatile substances, such as salts and proteins, are mainly expired in small droplets, and further diluted with exhaled water vapours [8,9]. It is thought that the droplets are formed as a result of random convective processes, and may not be directly related to water vapour production.
This has raised the question of variable droplet dilution, and given rise to some concerns regarding the interpretation of EBC biomarkers on the basis of their absolute concentration [9]. Some authors have suggested normalising for ion concentrations (Na+, Cl-, K+) or urea, assuming that they are equally concentrated in the airway lining fluid and serum of healthy and diseased subjects [8,10], and conductivity measurements of lyophilised EBC have also been proposed as a normalization factor [11]. On the other hand, the use of non-volatile parameters to normalise the volatile or semi-volatile compounds in EBC could ignore their condensation pathways as their ex vivo evaporation from the airways is different from droplet condensation in EBC collection systems. Ideally, a normalisation factor should be identified among substances with the same physical and chemical characteristics (i.e., volatility and solubility) as the measured parameters.
Volatility is generally assessed by means of a status diagram, which represents the relationship between the compounds' vapour pressure and temperature [12]. In a condensing system, the vapour pressure of volatile compounds depends on condensation temperature, and so the physical and chemical properties of exhaled compounds can be weighted by changing the condensation temperature. This may be particularly relevant in the case of EBC, in which the condensation of each substance may be affected by the presence of other compounds in the same solution.
A new type of condenser has been specifically designed to control the temperature of EBC collection, and test the effect of different condensation temperatures on the recovery of selected biomarkers (hydrogen peroxide, malondialdehyde). We also measured the conductivity after EBC lyophilisation, a parameter that reflects the overall concentration of salts. A rigorously precise study design controlling these aspects may represent a major advance in our understanding of condensation mechanisms, and in the validation of EBC as a suitable source of biomarkers reflecting the pathobiology underlying lung diseases.
Methods
Subjects
EBC was collected from 24 healthy non-smokers (table 1): i.e. asymptomatic and non-atopic individuals with normal spirometry results who showed no bronchial hyper-responsiveness to methacholine. The study was conducted in conformity with the declaration of Helsinki and was approved by the Ethical Committee of the University of Parma. All of the subjects gave their informed written consent.
Table 1 Characteristics of the study subjects. Data are expressed as mean ± SD.
Number of Subjects 24 (13 M; 11 F)
Age, Years 30 ± 4
Smokers/Ex-Smokers/Non Smokers 0/3/21
FVC, % of predicted 108.4 ± 10.4
FEV1, % of predicted 103.7 ± 11.4
FEV1/FVC, % of predicted 81.9 ± 6.3
Study design
The number of volunteers was exactly the same as the total number of possible sequence combinations of 4 temperatures (4! = 24). EBC was randomly collected from each subject using a different sequence of the four temperatures. The duration of the clinical part of the study was about two weeks in order to avoid the possible influence of climatic changes on EBC parameters. In order to assess inter-session variability, 10 of the 24 subjects underwent four consecutive EBC collections at the two extremes of temperature (-10°C and +5°C).
Collecting system
TURBO-DECCS (Transportable Unit for Research on Biomarkers Obtained from Disposable Exhaled Condensate Collection Systems) is a new and commercially available device for EBC collection (ItalChill, Parma, Italy). TURBO is a refrigerating system relying on a thermo electrical module giving rise to Peltier effect, which is activated by an electronic circuit in continuous current fed with main voltages. The cold side of the Peltier module is connected to an aluminium support shaped to house the test tube. A thermostat allows the collecting temperature to be regulated with a tolerance of ±1°C. The working temperature is adjustable from -10°C to room temperature or higher.
TURBO is supplied with DECCS, a disposable polyethylene device for collecting EBC that consists of a mouthpiece equipped with a one-way valve and saliva trap, connected to a collecting vial (50 ml) by means of a tube.
EBC collection
The subjects were asked to breath tidally through the mouthpiece without a nose clip for 10 minutes at collecting temperatures of -10°C, -5°C, 0°C and +5°C, with a 5-min interval after every collection. The subjects collected their EBC all in the same laboratory and no one collected EBC at home. EBC from every subject was collected according to different temporal combinations of the four temperatures covering the 24 possible combinations. The subjects were strictly instructed to maintain constant tidal breathing during the test and to form a complete seal around the mouthpiece; excess saliva was periodically eliminated and the mouth was rinsed with water. Salivary contamination was excluded by means of the colorimetric detection of alpha-amylase (Infinity amylase reagent, Sigma, Milan, Italy). The EBC samples were centrifuged for 1 min at 1000 g immediately after collection so that all of the water droplets were driven to the bottom of the flask, and the total volumes were measured. The samples were then stored at -80°C until analysis.
EBC analysis
Volume
The collected EBC volume was measured using a calibrated 200 μl pipette (Gilson International, Den Haag, The Netherlands) with an experimental error of ±10 μl.
Conductivity
EBC was lyophilised using a Heto FD 1.0 lyophilizer (Heto, Allerod, Denmark) and then re-suspended in ultra-pure water (Sigma, St. Louis, MO, USA). Conductivity was measured by means of an Istek 430c conductivity meter (Istek, Seoul, Republic of Korea). The Limit of Detection (LOD) of the method was about 0.1 μS/cm. Contamination of the storage vials was carefully assessed and excluded.
Hydrogen peroxide (H2O2)
H2O2 in EBC was measured as previously described [13] using a commercial kit (Amplex Red Hydrogen Peroxide assay kit, Molecular Probes, Eugene, USA) with a LOD of 0.01 μM. The analysis was made within 2–3 days of EBC collection after freezing at -80°C, in order to avoid the relative instability of H2O2 in EBC [14].
Malondialdehyde (MDA)
EBC MDA was measured by means of liquid chromatography-mass spectrometry tandem (LC-MS/MS) as previously described [15] within two weeks of EBC collection after storage at -80°C because of the relative high stability of MDA [16].
Statistical analysis
Data distribution was assessed using the Shapiro-Wilk test. Mean values ± SD were used for the normally distributed data, and geometric means [geometric SD] for the data with a lognormal distribution. Between-group differences were calculated using one-way ANOVA for repeated measures, followed by Tuckey's post-hoc test using the experimental data points or their logarithms, depending on the distribution of the experimental values. Regressions were performed using the least-squares method on either the experimental data points or their logarithms, using the Pearson's correlation coefficient to test goodness-of-fit. Intra-individual variability due to repeated measures was assessed using the dummy variables method for the multiple regression analysis [17]. A significance level of 0.05 was chosen for all of the statistical tests. The data were statistically analysed using two software programmes: SPSS 12.0 (SPSS inc., Chicago, IL, USA) and PRISM 3.0 (Graphpad Sofware, San Diego, CA, USA).
Results
Table 2 shows the inter-session variability in 10 of the 24 subjects at the two extreme temperatures (-10°C and +5°C). No statistically significant differences were found in the total volume, H2O2 level or conductivity values of the four consecutive collections.
Table 2 Inter-session variability of the selected parameters. t1, ...., t4 represent four consecutive EBC collections from 10 healthy subjects. No significant differences were found using repeated measures ANOVA followed by Tuckey's post-hoc test. MDA was excluded from the analysis because its reliability has been previously measured [15,16]. Mean ± SD or geometric mean [geometric SD] are also reported.
Total Volume (μl) H2O2 (μM) Conductivity (μS/cm)
T t1 t2 t3 t4 t1 t2 t3 t4 t1 t2 t3 t4
-10°C 1080 ± 270 1050 ± 220 1120 ± 250 1150 ± 220 0.097 [1.85] 0.092 [1.95] 0.108 [1.97] 0.112 [1.95] 1.8 [2.26] 2.6 [2.76] 2.4 [2.75] 1.5 [2.19]
+5°C 610 ± 200 680 ± 230 660 ± 250 600 ± 250 0.135 [1.92] 0.118 [1.88] 0.125 [1.91] 0.140 [2.00] 6 [2.95] 4.8 [2.70] 5.5 [2.50] 6.5 [2.33]
The measurements of the variables in the samples collected at different temperatures are shown in Figures 1, 2, 3, 4, which also show their distribution parameters (mean ± SD or geometric mean [geometric SD]). There was a clear trend toward increasing EBC sample volumes with decreasing collection temperatures (Figure 1), which affected both the concentration and absolute amounts of the selected analytes to different extents.
Figure 1 Variation in total EBC volume collected at different cooling temperatures. The horizontal grey lines indicate mean values. The parameters of the distributions (mean ± SD) at different temperatures and the significant differences are also reported. * = p < 0.05; ** = p < 0.01.
Figure 2 Concentration (a) and absolute amount (b) of EBC H2O2 at different cooling temperatures. The horizontal grey lines indicate geometric mean (a) and mean values (b). The parameters of the distributions (geometric mean [geometric SD] and mean ± SD for a and b respectively) at different temperatures and the significant differences are also reported. * = p < 0.05; ** = p < 0.01.
Figure 3 Concentration (a) and absolute quantity (b) of EBC MDA at different cooling temperatures. The horizontal grey lines indicate geometric mean (a) and mean values (b). The parameters of the distributions (geometric mean [geometric SD] and mean ± SD for a and b respectively) at different temperatures and the significant differences are also reported. * = p < 0.05; ** = p < 0.01.
Figure 4 EBC conductivity (a) and EBC conductivity*EBC volume values (b) at different cooling temperatures. The horizontal grey lines indicate geometric mean values. The parameters of the distributions (geometric mean [geometric SD]) at different temperatures and the significant differences are also reported. * = p < 0.05; ** = p < 0.01.
All of the differences in volume between temperature pairs were statistically significant, except for the comparison between -5°C and 0°C. At a fixed temperature, the total volume of expired air closely correlated with the total condensed volume (r > 0.95, data not shown).
The H2O2 concentrations measured in the EBC samples collected at different cooling temperatures (Figure 2a) were significantly different between -10°C and +5°C (p < 0.05). The absolute amount of H2O2 (in pmol) was calculated by multiplying its concentration by the corresponding total condensed volume. As shown in Figure 2b, a progressive increase in the amount of EBC H2O2 was recorded as the cooling temperature decreased, with significant differences between -10 and +5°C and between -10 and 0°C (p < 0.01).
Figure 3a shows the differences in MDA concentration measured in the EBC samples collected at different cooling temperatures (between -10 and 0°C and between -5 and +5°C, p < 0.05; between -10 and +5°C, p < 0.01). Figure 3b shows that the absolute amount of MDA (in pmol) decreased with increasing temperature, with some residual statistically significant differences between -10 and 0°C (p < 0.05) and between -10 and +5°C, (p < 0.01).
There was a progressive increase in EBC conductivity values at different cooling temperatures (Figure 4a), with significant differences between -10 and 0°C, between -5 and +5°C (p < 0.05) and between -10 and +5°C (p < 0.01). These differences disappeared when conductivity was multiplied by volume (thus expressing results as μS*cm2), as shown Figure 4b.
On the basis of the geometric standard deviation (GSD) of the values shown in Figures 2a, 3a and 4a, conductivity was more variable (range 2.17–3.35) than either H2O2 (range 1.81–1.96) or MDA (range 1.29–1.37) concentrations.
Table 3 shows the parameters of regressions between measured variables: Pearson's correlation coefficient with its significance, and the intercept with Y axis (A) and the slope (B). The regressions were performed using the least-square method either on experimental data points or their logarithms, with Pearson's correlation coefficient being used to test the goodness-of-fit. Intra-individual variability due to repeated measures was assessed using the dummy variable method for multiple regression analysis. The results indicated a weak negative correlation between total volume and H2O2 (r = -0.29, p < 0.01) but, considering B values in table 3, the intra-individual contribution was not significant. There was also a negative correlation between total volume and MDA (r = -0.54, p < 0.01) with a similar inter-individual and intra-individual contribution. Conductivity showed a similar, albeit weaker correlation (r = -0.34, p < 0.01). H2O2 and MDA levels positively correlated (r = 0.53, p < 0.01). There was no significant correlation between H2O2 and conductivity (r = 0.1, p = ns), but a significant, albeit weak correlation between MDA and conductivity (r = 0.21, p < 0.05), with a higher intra-individual contribution to variability.
Table 3 Regressions between the measured variables. Regressions between the measured variables using the general model (second column) and the dummy variable method (third column). ns = not significant. Cond. = Conductivity, Vol. = Volume. r = Pearson's correlation coefficient; p = significance of r; A: intercept with Y-axis (±SD is also expressed); B: Slope of the regression line (±SD is also expressed); p (B): significance of B. The correlations refer to the concentrations of H2O2 and MDA. Ei = dummy variable. Its value is 1 for the i-th subject, -1 for the last subject, 0 elsewhere.
H2O2 vs Vol. Log (H2O2) = A + B*Volume Log (H2O2) = A + B*Volume + ΣiCiEi
r P A B p (B) A B p (B)
-0.29 <0.01 -0.73 ± 0.08 (-2.7 ± 0.9)*10-5 <0.01 -0.88 ± 0.08 (-8.5 ± 5.2)*10-5 ns
MDA vs Vol. Log (MDA) = A + B*Volume Log (MDA) = A + B*Volume + ΣiCiEi
r P A B p (B) A B p (B)
-0.54 <0.01 1.01 ± 0.04 (-2.4 ± 0.4)*10-4 <0.01 1.00 ± 0.04 (-2.3 ± 0.4)*10-4 <0.01
Cond. vs Vol. Log (Cond.) = A + B*Volume Log (Cond.) = A + B*Volume + ΣiCiEi
r P A B p (B) A B p (B)
-0.34 <0.01 0.94 ± 0.13 (-5.2 ± 1.5)*10-4 <0.01 1.05 ± 0.13 (-6.5 ± 1.4)*10-4 <0.01
MDA vs H2O2 Log (MDA) = A + B*Log (H2O2) Log (MDA) = A + B*Log (H2O2) + ΣiCiEi
r P A B p (B) A B p (B)
0.53 <0.01 1.05 ± 0.04 0.25 ± 0.04 <0.01 1.16 ± 0.09 0.37 ± 0.10 <0.01
H2O2 vs Cond. Log (H2O2) = A + B*Log (Cond.) Log (H2O2) = A + B*Log (Cond.) + ΣiCiEi
r P A B p (B) A B p (B)
0.12 Ns Ns ns
MDA vs. Cond. Log (MDA) = A + B*Log (Cond.) Log (MDA) = A + B*Log (Cond.) + ΣiCiEi
r P A B p (B) A B p (B)
0.21 <0.05 0.78 ± 0.02 0.06 ± 0.03 <0.05 0.76 ± 0.02 0.10 ± 0.03 <0.01
Discussion
Although EBC is mostly water, it contains appreciable concentrations of volatile and non-volatile solutes. The presence of salts and peptides in EBC suggests a transfer of non-volatile compounds to the air phase, probably in the form of small droplets to allow the vapour stream to go through convective processes. Mathematical models based on in vitro experiments [18] have been developed in order to understand more about the physical phenomenon of droplet formation, and the size distribution of exhaled droplets has been characterized [19,20]. Other Authors have proposed complex mathematical models designed to account for the presence of non-volatile solutes in ambient air [21,22].
Various approaches have been proposed as a mean of normalising biomarker concentrations in EBC [8-11], including the use of endogenous non-volatile substances (ions, urea) or parameters (osmolality, conductivity after lyophilisation). One controversial assumption is that serum and airway surface lining fluid are isotonic, but this is not supported by available evidence [23-25]. Although a normalisation factor may be useful for non-volatile molecules such as proteins and electrolytes, the ex vivo volatility of some currently measured biomarkers in EBC is still unclear, and their normalization by any non-volatile factor could actually lead to ignore their different exhalation pathways and consequent focusing ability in EBC collecting systems.
In this study, we assessed the ex vivo volatility of H2O2 and MDA, which are respectively considered to be reliable biomarkers of airway inflammation and membrane peroxidation [26,27]; the volatility and solubility of H2O2 are well known in aqueous solutions [28], but little is known about MDA. We also measured total condensed volume, which reflects overall subject ventilation [29], and conductivity after EBC lyophilisation, which reflects the concentration and charge of non-volatile electrolytes [11].
As expected, total condensed volume inversely related to temperature. On a pretty constant subject expiration, the number of condensed water molecules derived from the aqueous vapour phase depends on condensation temperature; with lower temperatures condensating a larger number of water molecules. Assuming an EBC density of approximately 1 g/cm3 (near to that of pure water), the increase in the number of water molecules in the selected range of cooling temperatures can be calculated using the equation:
% increase water = Volume (-10°C)/Volume (+5°C) = 1.72,
On the basis of the mean values calculated in Figure 1. This ratio is significantly different from 1 (p < 0.01).
The same equation (with different means and geometric means depending on the data distribution) was used to calculate the absolute amount of H2O2 and MDA. As the ratios were significantly different from 1 (1.37 for H2O2 and 1.28 for MDA, p < 0.01, Figures 2b and 3b), these biomarkers can be considered volatile compounds ex vivo. In both cases, the percentage of recovery is less than that of water. Therefore, as the condensation temperature decreased, the increase in their absolute amount was less than that of water.
The behaviour of EBC conductivity is an important comparative parameter. Like H2O2 and MDA, conductivity increased with temperature, although in this case the absolute increase was greater and inversely related to volume (Figure 4a). Assuming that EBC conductivity is proportional to mono-charged ion EBC concentration, which should represent more than 80% of the total [11], the product of conductivity by volume should be approximately proportional to the number of electrolyte molecules and, as expected, the differences between collection temperatures were no longer significant (Figure 4b). This may be due to the fact that the number of expired droplets (and therefore the absolute amount of recovered ions) is not temperature dependent, whereas the observed decrease in conductivity at lower temperatures should only depend on number of condensed water molecules with the subsequent dilution of non-volatile substances. We calculated the percent increase in salt molecules using the equation:
% increase salts molecules ≈ Conductivity*Volume (-10°C)/Conductivity*Volume (+5°C) = 0.66,
on the basis of the geometric mean values shown in figure 4b. The ratio is not significantly different from 1.
On the basis of our data, H2O2 and MDA should both be considered volatile substances ex vivo, whereas conductivity should not. The normalisation of these volatile substances by non-volatile compounds or parameters would therefore not take into account their specific excretion pathway. Further, the high variability of conductivity (Figure 4a) suggests that its level in EBC is not regulated by a specific biological elimination mechanism, but depends on a random process, that would make its use as a normalisation factor hazardous. Effros et al. [9] suggested that volatile compounds should be measured in the gas phase of expiration rather than in EBC, but our findings show that, although the determination of low-volatility compounds is influenced by the condensation temperature, their determination in EBC is not nonsensical provided that the EBC collection temperature is fixed: at the relatively high condensation temperature of +5°C, about 70% of the total molecules measured at -10°C were present in solution.
The intra-session variability study (Table 2) showed that temperature-related differences in collected marker levels can not be attributed to intra-individual variability. MDA reliability was not assessed as it has already been demonstrated [15,16].
The between-variable correlations and regressions gave further information concerning the mutual relationships of the compounds. When regression was calculated without making any distinction for repeated measures, the contribution of temperature was mixed with other possible contributions, such as that of ventilatory volume, which closely correlated with the total collected volume at a fixed temperature. Isolation of the intra-individual effect by means of the dummy variable method distinguished the effect of temperature from the other contributions as the ventilatory volume of the same subject was kept constant at the different temperatures, and made it possible to estimate its effect on the parameters describing the regression.
EBC H2O2 concentrations showed a weak negative correlation with EBC total volume (r = -0.29), but this was not significant when only the intra-individual effect was considered (Table 3). Although the temperature-induced variations in volume did not correlate with the variations in H2O2 concentration, the inter-individual variability in volume due to other contributions moderately affected H2O2 concentrations. On the contrary, EBC MDA concentration negatively correlated with total EBC volume (r = -0.54), with a similar contribution of inter-individual and intra-individual effect (Table 3). As a result, total volume could be a relevant parameter, particularly if the EBC biomarker correlates with EBC volume and the effect of condensation temperature is ruled out. When sampling occurs at a constant temperature, any inter-individual difference in EBC volume can be ascribed to total ventilation volume. Under these circumstances, the use of total EBC volume as a covariate could normalise the effect of total ventilation volume on biomarker concentrations.
H2O2 and MDA positively correlated (r = 0.53). These data suggest that, in addition to their volatility, hydrogen peroxide production and lipid peroxidation could also be related processes in vivo in healthy subjects, as already observed in subjects affected by respiratory tract inflammation [30].
Finally, although there was no correlation between H2O2 concentration and conductivity values, thus reinforcing the idea that non-volatile compounds and H2O2 in EBC have different physico-chemical properties, a very weak positive correlation was found between MDA and conductivity (r = 0.21): this suggests that a slight contribution to the total concentration of MDA in EBC could derive from MDA-containing droplets. In fact, the lack of correlation between conductivity and volatile components indicates that non-volatile ions reflect the number of airway lining fluid droplets joining the vapour stream, a mechanism that would complement evaporation. The latter seems to be the main determinant of H2O2 (which is volatile) and MDA (which is also slightly volatile) content in EBC.
Conclusion
On the basis of the present study, we suggest that:
1. the temperature of EBC collection should be controlled and reported;
2. cooling temperatures should be chosen on the basis of analytical needs (required EBC volumes, sensitivity of the method, etc.);
3. water is the main variable dilution factor, and so total condensed volume should be recorded;
4. the cooling temperature related to EBC collection may differently influence biomarkers and normalizing factors, which should belong to the same class as the analytes requiring normalization (e.g., in terms of relative volatility and solubility).
Abbreviations
EBC = Exhaled Breath Condensate; MDA = Malondialdehyde; LOD = Limit of Detection; LC-MS/MS = liquid chromatography-mass spectrometry tandem; TURBO-DECCS = Transportable Unit for Research on Biomarkers Obtained from Disposable Exhaled Condensate Collection Systems.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MG: substantial contribution to conception and design, acquisition of data, analysis and interpretation of data, involved in drafting the article.
AC: substantial contribution to conception and design, collection of samples, revision of the draft critically for important intellectual content.
RA:, acquisition of data, revision of the draft critically for important intellectual content.
DP: substantial contribution to conception and design, revision of the draft critically for important intellectual content.
PM: substantial contribution to conception and design, revision of the draft critically for important intellectual content.
MVV: substantial contribution to conception and design, acquisition of data, revision of the draft critically for important intellectual content.
MC: substantial contribution to conception and design, analysis and interpretation of data, involved in drafting the article.
AM: substantial contribution to conception and design, statistical analysis and interpretation of data, involved in drafting the article, final approval of the version to be published.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by the National Heart, Blood and Lung Institute (NHLBI), Bethesda, MD, USA (grant 1R01 HL72323-01), and by the Italian Ministry of Education, University and Research (PRIN 200306145). Contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the NHLBI or the National Institutes of Health.
==== Refs
Mutlu GM Garey KW Robbins RA Danziger LH Rubinstein I Collection and analysis of exhaled breath condensate in humans Am J Respir Crit Care Med 2001 164 731 737 11549524
Mondino C Ciabattoni G Koch P Pistelli R Trove A Barnes PJ Montuschi P Effects of inhaled corticosteroids on exhaled leukotrienes and prostanoids in asthmatic children J Allergy Clin Immunol 2004 114 761 767 15480313 10.1016/j.jaci.2004.06.054
Effros RM Su J Casaburi R Shaker R Biller J Dunning M Utility of exhaled breath condensates in chronic obstructive pulmonary disease: a critical review Curr Opin Pulm Med 2005 11 135 139 15699785 10.1097/00063198-200503000-00006
Carpagnano GE Foschino-Barbaro MP Resta O Gramiccioni E Carpagnano F Endothelin-1 is increased in the breath condensate of patients with non-small-cell lung cancer Oncology 2004 66 180 184 15218307 10.1159/000077992
Carpagnano GE Kharitonov SA Wells AU Pantelidis P Du Bois RM Barnes PJ Increased vitronectin and endothelin-1 in the breath condensate of patients with fibrosing lung disease Respiration 2003 70 154 160 12740512 10.1159/000070062
Gessner C Hammerschmidt S Kuhn H Seyfarth HJ Sack U Engelmann L Schauer J Wirtz Exhaled breath condensate acidification in acute lung injury Respir Med 2003 97 1188 1194 14635972 10.1016/S0954-6111(03)00225-7
Goldoni M Catalani S De Palma G Manini P Acampa O Corradi M Bergonzi R Apostoli P Mutti A Exhaled breath condensate as a suitable matrix to assess lung dose and effects in workers exposed to cobalt and tungsten Environ Health Perspect 2004 112 1293 1298 15345342
Dwyer TM Sampling airway surface liquid: non-volatiles in the exhaled breath condensate Lung 2004 182 241 250 15636196 10.1007/s00408-004-2506-3
Effros RM Dunning MB 3rdBiller J Shaker R The promise and perils of exhaled breath condensates Am J Physiol Lung Cell Mol Physiol 2004 287 L1073 80 15531756 10.1152/ajplung.00069.2004
Effros RM Hoagland KW Bosbous M Castillo D Foss B Dunning M Gare M Lin W Sun F Dilution of respiratory solutes in exhaled condensates Am J Respir Crit Care Med 2002 165 663 669 11874811
Effros RM Biller J Foss B Hoagland K Dunning MB Castillo D Bosbous M Sun F Shaker R A simple method for estimating respiratory solute dilution in exhaled breath condensates Am J Respir Crit Care Med 2003 168 1500 1505 14512268 10.1164/rccm.200307-920OC
Atkins PW Physical Chemistry 1998 6 Oxford: Oxford University Press
Corradi M Alinovi R Goldoni M Vettori MV Folesani G Mozzoni P Cavazzini S Bergamaschi E Rossi L Mutti A Biomarkers of oxidative stress after controlled human exposure to ozone Toxicol Lett 2002 134 219 225 12191881 10.1016/S0378-4274(02)00169-8
Nowak D Kalucka S Bialasiewicz P Krol M Exhalation of H2O2 AND Thiobarbituric acid reactive substances (TBARS) by healthy subjects Free Rad Biol Med 2001 30 178 186 11163535 10.1016/S0891-5849(00)00457-3
Andreoli R Manini P Corradi M Mutti A Niessen WM Determination of patterns of biologically relevant aldehydes in exhaled breath condensate of healthy subjects by liquid chromatography/atmospheric chemical ionization tandem mass spectrometry Rapid Commun Mass Spectrom 2003 17 637 645 12661015 10.1002/rcm.960
Corradi M Pignatti P Manini P Andreoli R Goldoni M Poppa M Moscato G Balbi B Mutti A Comparison between exhaled and sputum oxidative stress biomarkers in chronic airway inflammation Eur Respir J 2004 24 1011 1017 15572547 10.1183/09031936.04.00002404
Glantz SA Slinker BK Primer of Applied Regression and Analysis of Variance 2000 2 Columbus, OH: McGraw-Hill Education
Moriarty JA Grotberg JB Flow-induced instabilities of a mucus-serous bilayer J Fluid Mech 1999 397 1 22 10.1017/S0022112099005704
Papineni RS Rosenthal FS The size distribution of droplets in the exhaled breath of healthy human subjects J Aerosol Med 1997 10 105 116 10168531
Edwards DA Man JC Brand P Katstra JP Sommerei K Stone HA Nardell E Scheuch G Inhaling to mitigate exhaled bioaerosols Proc Natl Acad Sci USA 2004 101 17383 17388 15583121 10.1073/pnas.0408159101
Mikhailov E Vlasenko S Niessner R Poschl U Interaction of aerosol particles composed of protein and salts with water vapor: hygroscopic growth and microstructural rearrangement Atmos Chem Phys Discuss 2003 3 4755 4832
Clement CF Mean field theory for condensation on aerosols and application to multi-component organic vapours Aerosol Sci 2003 34 27 40 10.1016/S0021-8502(02)00154-4
Joris L Dab I Quinton PM Elemental composition of human airway surface fluid in healthy and diseased airways Am Rev Respir Dis 1993 148 1633 1637 8256912
Knowles MR Robinson JM Wood RE Pue CA Mentz WM Wager GC Gatzy JT Boucher RC Ion Composition of Airway Surface Liquid of Patients with Cystic Fibrosis as Compared with Normal and Disease-control Subjects J Clin Invest 1997 100 2588 2595 9366574
Widdicombe JH Altered NaCl concentration of airway surface liquid in cystic fibrosis Pflugers Arch 2001 443 S8 10 11845295
De Benedetto F Aceto A Dragani B Spacone A Formisano S Pela R Donner CF Sanguinetti CM Long-term oral n-acetylcysteine reduces exhaled hydrogen peroxide in stable COPD Pulm Pharmacol Ther 2005 18 41 47 15607126 10.1016/j.pupt.2004.09.030
Corradi M Rubinstein I Andreoli R Manini P Caglieri A Poli D Alinovi R Mutti A Aldehydes in exhaled breath condensate of patients with chronic obstructive pulmonary disease Am J Respir Crit Care Med 2003 167 1380 1386 12522029 10.1164/rccm.200210-1253OC
Introduction to Hydrogen Peroxide-Physical and Chemical Properties – Physical
McCafferty JB Bradshaw TA Tate S Greening AP Innes JA Effects of breathing pattern and inspired air conditions on breath condensate volume, pH, nitrite, and protein concentrations Thorax 2004 59 694 698 15282391 10.1136/thx.2003.016949
Mylonas C Kouretas D Lipid peroxidation and tissue damage In Vivo 1999 13 295 309 10459507
|
16137323
|
PMC1236937
|
CC BY
|
2021-01-04 16:30:12
|
no
|
BMC Pulm Med. 2005 Sep 1; 5:10
|
utf-8
|
BMC Pulm Med
| 2,005 |
10.1186/1471-2466-5-10
|
oa_comm
|
==== Front
BMC Pulm MedBMC Pulmonary Medicine1471-2466BioMed Central London 1471-2466-5-121615939810.1186/1471-2466-5-12Research ArticleThe impact of prior outpatient ACE inhibitor use on 30-day mortality for patients hospitalized with community-acquired pneumonia Mortensen Eric M [email protected] Marcos I [email protected] Antonio [email protected] Jacqueline [email protected] VERDICT Research Center, Audie L Murphy VA Hospital, San Antonio, Texas, USA2 Division of General Medicine, The University of Texas Health Science Center at San Antonio, USA3 Division of Pulmonary and Critical Care Medicine, The University of Texas Health Science Center at San Antonio, USA2005 13 9 2005 5 12 12 6 4 2005 13 9 2005 Copyright © 2005 Mortensen et al; licensee BioMed Central Ltd.2005Mortensen 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 studies suggest that angiotensin-converting enzyme (ACE) inhibitors may have beneficial effects for patients at risk for some types of infections. We examined the effect of prior outpatient use of ACE inhibitors on mortality for patients hospitalized with community-acquired pneumonia.
Methods
A retrospective cohort study conducted at two tertiary teaching hospitals. Eligible subjects were admitted with a diagnosis of, had a chest x-ray consistent with, and had a discharge ICD-9 diagnosis of pneumonia. Subjects were excluded if they were "comfort measures only" or transferred from another acute care hospital. Subjects were considered to be on a medication if they were taking it at the time of presentation.
Results
Data was abstracted on 787 subjects at the two hospitals. Mortality was 9.2% at 30-days and 13.6% at 90-days. At presentation 52% of subjects were low risk, 34% were moderate risk, and 14% were high risk. In the multivariable conditional logistic regression analysis, after adjusting for potential confounders, the use of ACE inhibitors at presentation (odds ratio 0.44, 95% confidence interval 0.22–0.89) was significantly associated with 30-day mortality.
Conclusion
Prior outpatient use of an ACE inhibitor was associated with decreased mortality in patients hospitalized with community-acquired pneumonia despite their use being associated with comorbid illnesses likely to contribute to increased mortality. Confirmatory studies are needed, as well as research to determine the mechanism(s) of this protective effect.
==== Body
Background
Community-acquired pneumonia is the seventh leading cause of death and the leading cause of infectious death in the United States [1]. Although mortality due to community-acquired pneumonia decreased significantly with the introduction of antibiotics in the 1950s since that time mortality has been stable or increasing [2]. Despite this, only a few new classes of antibiotics have been added to the armamentarium for treating community-acquired pneumonia in the last 20 years and no new classes of medications beyond antibiotics have been added since the 1950s.
Recent studies have demonstrated that elderly patients taking ACE inhibitors have significantly decreased rates of aspiration pneumonia and may have decreased mortality [3-5]. The hypothesized reason for this protective effect is an increase in substance P which accentuates the cough reflex. Other studies have demonstrated significant changes in systemic cytokine levels for subjects treated with ACE inhibitors [6-8]. These cytokines play an important role in host defense mechanisms for patients with community-acquired pneumonia but under certain conditions may lead to septic shock or acute respiratory distress syndrome (ARDS) [9-11].
The study aim was to assess the effects of prior outpatient ACE inhibitor use on 30-day mortality for patients hospitalized with community-acquired pneumonia.
Methods
This a retrospective cohort study of patients hospitalized with community-acquired pneumonia at 2 academic tertiary care hospitals in San Antonio, Texas. Both hospitals are teaching affiliates of the University of Texas Health Science Center at San Antonio. The Institutional Review Board of the University of Texas Health Science Center at San Antonio approved the research protocol with exempt status.
Study sites/inclusion and exclusion criteria
We identified all patients admitted to the study hospitals between January 1, 1999 and December 1, 2002 with a primary discharge diagnosis of pneumonia (ICD-9 codes 480.0–483.99 or 485–487.0) or secondary discharge diagnosis of pneumonia with a primary diagnosis of respiratory failure (518.81) or sepsis (038.xx). Subjects were included if they were 1) greater than 18 years of age, 2) had an admission diagnosis of community-acquired pneumonia documented in the medical record, and 3) had a radiographically confirmed infiltrate or other finding consistent with community-acquired pneumonia on chest x-ray or CT obtained within 24 hours of admission.
Exclusion criteria included 1) having been discharged from an acute care facility within 14 days of admission, 2) transfer after being admitted to another acute care hospital, and 3) being comfort measures only on this admission. If a subject was admitted more than once during the study period, only the first hospitalization was abstracted
Data abstraction
Chart review data included: demographics, comorbid conditions, physical examination findings, laboratory data, and chest radiograph reports. In addition, data on important processes of care measures for patients hospitalized with community-acquired pneumonia were also abstracted: first dose of antibiotics within 8 hours of admission, collection of blood cultures prior to antibiotic administration, and obtaining blood cultures and oxygen saturation measurement within 24 hours of presentation [12]. Antimicrobial therapy was considered guideline-concordant if it agreed with either the 2000 Infectious Diseases Society of America or 2001 American Thoracic Society guidelines [13,14]. Information on all outpatient medications that were either 1) reported as currently being taken by the patient at presentation, or 2) listed in the electronic medical record, were recorded.
Mortality was assessed using information from the Texas Department of Health and Department of Veteran Affairs clinical database. Mortality status was assessed through December 31, 2002.
Risk adjustment
The pneumonia severity index was used to assess severity of illness at presentation [15]. The pneumonia severity index is a validated prediction rule for 30-day mortality in patients with community-acquired pneumonia. This rule is based on three demographic characteristics, five comorbid illnesses, five physical examination findings, and seven laboratory and radiographic findings from the time of presentation. Patients are classified into five risk classes with 30-day mortality ranging from 0.1% for class I to 27% for class V for patients enrolled in the PORT cohort study [15].
Outcome
We used 30-day mortality as the outcome for this study. Previous research has demonstrated that 30-day mortality is primarily due to the community-acquired pneumonia rather than other co-existing co-morbid conditions [16,17]. Therefore by using 30-day mortality as our outcome we are able to minimize the effect of the ACE inhibitor use on other co-morbid conditions.
Statistical analyses
Univariate statistics were used to test the association of sociodemographic and clinical characteristics with all-cause 30-day mortality. Categorical variables were analyzed using the Chi-square test and continuous variables were analyzed using Student's t-test.
A propensity score technique was used to balance covariates associated with ACE inhibitor use between groups [18,19]. The use of the propensity score technique provides a way, in non-randomized studies, to control for pretreatment differences by defining sets of comparable patients. The propensity score was derived from a logistic regression model, and a dichotomous indicator variable indexing whether a patient was on an ACE inhibitor was our dependent variable. The covariates used to develop the propensity score included the pneumonia severity index (which includes comorbid conditions such as congestive heart failure and chronic renal insufficiency), history of hypertension, and history of diabetes mellitus.
A multivariable conditional logistic regression model was derived with 30-day mortality as the dependent variable, and the propensity score as the matching variable [20]. The independent variables in the model were the use of ACE inhibitor at presentation, and process of care measures (initial antibiotics within 8 hours and obtaining blood cultures prior to initial dose of antibiotics, and whether antimicrobial therapy was guideline concordant). Interactions were assessed using cross-product terms between the medications and all of the other variables retained in the models. No significant interactions terms were noted, so they were excluded from the final models. All analyses were performed using STATA version 8 (Stata Corporation, College Station, Texas).
Results
Data was abstracted on 787 patients at the two hospitals. The mean age was 60 years with a standard deviation of 16 years. The population was 79% male, 84% were admitted through the emergency department, and 20% were admitted to the intensive care unit within the first 24 hours after admission. Mortality was 9.2% at 30-days and 13.6% at 90-days. By pneumonia severity index, 52% were low risk (pneumonia severity index classes I-III), 34% were moderate risk (pneumonia severity index class IV), and 14% were high risk (pneumonia severity index class V). Regarding community-acquired pneumonia-related processes of care, 28% received the initial dose of antibiotics within 4 hours of presentation and an additional 22% received the initial antibiotic dose within 8 hours, 76% of patients had blood cultures obtained within 24 hours and prior to antibiotics, and oxygenation was assessed at presentation in 91%.
Table 1 shows the demographic factors, clinical characteristics, and processes of care data for this population by 30-day mortality. In the univariate analysis numerous individual components of the pneumonia severity index were significantly associated with 30-day mortality including age, nursing home residency, history of congestive heart failure, history of malignancy, altered mental status, systolic blood pressure < 90 mmHg, tachycardia> 125 beats per minute, arterial acidosis, elevated blood urea nitrogen 30 mg/dl, serum sodium < 130 meq/l, and pleural effusion on chest radiograph. The only processes of care that were statistically significant were the assessment of oxygenation within 24 hours or presentation and use of guideline-concordant antibiotics. ACE inhibitor use was significantly associated (p = 0.05) with 30-day mortality.
Table 1 Subject Demographic and Clinical Characteristics by 30-Day Mortality*
30-Day Mortality
Variable Alive (n = 714) Dead (n = 72) p-value
Age, years +/- standard deviation 60.2+/-16.4 62.9 +/-16.4 0.09
Men 561 (79) 60 (83) 0.3
Nursing home resident 41 (6) 13 (18) <0.001
Emergency department admission 598 (84) 58 (81) 0.5
Admitted to intensive care ≤ 24 hours 118 (17) 36 (50) <0.001
Preexisting Comorbid Conditions
Congestive heart failure 105 (15) 18 (25) 0.02
Chronic pulmonary disease 195 (27) 23 (31) 0.4
History of stroke 93 (13) 12 (17) 0.4
Chronic liver disease 83 (12) 11 (15) 0.4
History of malignancy 58 (8) 20 (28) <0.001
Renal insufficiency 74 (10) 13 (18) 0.05
History, Physical, Laboratory, and Radiographic Data
Altered mental status 68(10) 17(24) <0.001
Respiratory rate > 30 per minute 71 (10) 11 (15) 0.2
Systolic blood pressure < 90 mmHg 16 (2) 5 (7) 0.02
Heart rate > 125 per minute 86 (12) 19 (26) 0.001
Temperature < 95° or > 104° 19 (3) 2 (3) 0.9
Arterial pH < 7.35 37 (5) 12 (17) <0.001
Arterial oxygenation < 90% 149 (21) 27 (38) 0.001
Hematocrit < 30% 64 (9) 8 (11) 0.5
Blood urea nitrogen > 30 mg/dL 135 (19) 33 (46) <0.001
Serum glucose > 250 mg/dL 71 (10) 5 (7) 0.4
Serum sodium < 130 meq/L 98 (14) 18 (25) 0.01
Pleural effusion 160 (11) 29 (35) 0.001
Pneumonia Severity Index
Class I-III 393 (54) 16 (22)
Class IV 240 (34) 26 (36)
Class V 82 (12) 30 (42) <0.001
Processes of Care
Initial antibiotics within 4 hours 201 (28) 22 (31) 0.7
Initial antibiotics within 8 hours 358 (50) 36 (50) 1.0
Blood cultures prior to antibiotics 540 (76) 55 (76) 0.9
Oxygenation assessed ≤ 24 hours 538 (75) 65 (90) 0.004
Guideline-concordant antibiotics 574 (80) 51 (71) 0.05
Outpatient Medications
ACE inhibitor 183 (25) 11 (15) 0.05
* Data are presented as number (%) or mean +/-standard deviation
Table 2 demonstrates clinical and demographic variables and the association with use/non-use of ACE inhibitors at presentation. Co-morbid conditions and demographic variables significantly associated with ACE inhibitor use included coronary artery disease, higher age, diabetes mellitus, chronic obstructive pulmonary disease, not currently smoking, congestive heart failure, prior stroke, no history of liver disease, and chronic renal disease. Physical exam and laboratory findings associated with ACE inhibitor use included decreased numbers of patients with heart rates > 125, higher rates of elevated blood urea nitrogen > 30 mg/dL, lower rates of serum hematocrit < 30%, and higher rates of serum glucose > 250 mg/dL. Of the 787 subjects 194 (25%) were on ACE inhibitors. There was greater ACE inhibitor use in those who were in the moderate (30%, n = 74) and severe (34%, n = 38) risk groups versus those with low severity of illness (19%, n = 77), p < 0.001.
Table 2 Use versus non-use of ACE inhibitors by demographic and clinical characteristics*
ACE Inhibitor Use
Variable Not on ACE inhibitor
(n= 593) On ACE inhibitor
(n=194) p-value
Age, years +/- standard deviation 58.2 +/-16.9 67.1+/-13.6 <0.001
Men 467 (79) 154(80) 0.8
Nursing home resident 43(7) 11(6) 0.4
Emergency department admission 498(84) 158(81) 0.38
Admitted to intensive care ≤ 24 hours 116(20) 38(20) 1.0
Preexisting Comorbid Conditions
Diabetes Mellitus 138(23) 92(47) <0.001
Coronary artery disease 132(22) 98(51) <0.001
Chronic pulmonary disease 151(25) 67(35) <0.001
Current tobacco use 200(31) 35(18) <0.001
Congestive heart failure 59(10) 64(33) <0.001
History of stroke 67(11) 38(20) 0.003
Chronic liver disease 82(14) 12(6) 0.004
History of malignancy 57(10) 21(11) 0.6
Renal insufficiency 54(9) 33(17) 0.002
History, Physical, Laboratory, and Radiographic Data
Altered mental status 62(10) 23(12) 0.6
Respiratory rate > 30 per minute 67(11) 15(8) 0.16
Systolic blood pressure < 90 mmHg 17(3) 4(2) 0.5
Heart rate > 125 per minute 87(15) 18(9) 0.06
Temperature < 95° or > 104° 18(3) 3(2) 0.26
Arterial pH < 7.35 33(6) 16(8) 0.18
Arterial oxygenation < 90% 128(22) 48(25) 0.4
Hematocrit < 30% 59(10) 13(7) 0.17
Blood urea nitrogen > 30 mg/dL 110(19) 58(30) <0.001
Serum glucose > 250 mg/dL 50(8) 26(13) 0.04
Serum sodium < 130 meq/L 95(16) 21(11) 0.08
Pleural effusion 141(24) 48(25) 0.8
Pneumonia Severity Index
Class I-III 332(56) 77(40)
Class IV 187(32) 79(41)
Class V 74(12) 38(20) <0.001
* Data are presented as number (%) or mean +/-standard deviation
In the multivariable conditional logistic regression analysis after adjusting for potential confounders the use of ACE inhibitors at presentation (odds ratio (OR) 0.44, 95% confidence interval (CI) 0.22–0.89) was associated with 30-day mortality. The other variables in the model included initial antibiotics within 4 hours (OR 1.09, 95% CI 0.64-1.85), obtaining blood cultures prior to antibiotics (OR 1.10, 95% CI 0.6-1.9), oxygenation assessment within 24 hours (OR 1.31, 95% CI 0.5–3.4), and use of guideline-concordant antimicrobial therapy (OR 0.6, 0.35–1.05).
Discussion
We found that prior outpatient use of ACE inhibitors was associated with decreased 30-day mortality for subjects hospitalized with community-acquired pneumonia. Our findings provide further support to other studies that demonstrate that ACE inhibitor use is associated with decreased mortality for patients with pneumonia.
There are several potential mechanisms for a beneficial effect on mortality for patients on ACE inhibitors. ACE inhibitors have been shown in clinical studies to increased serum levels of substance P, which is hypothesized to lead to a better gag reflex and increased clearance of secretions [3-5]. Our study specifically excluded patients with a discharge diagnosis of aspiration pneumonia; however, this diagnosis can be clinically difficult to distinguish from community-acquired pneumonia. Other studies have demonstrated that ACE inhibitors have significant immunomodulatory effects on circulating cytokines [6-8]. Therefore the beneficial effect may be due to blunting of the cytokine response so that these patients have lower rates of sepsis and/or ARDS.
Although our study was retrospective and subject to the recognized limitations of such studies, we carefully assembled our cohort from complete patient discharge data to avoid ascertainment bias. Additionally, during chart abstraction we encountered a very small amount (<5%) of missing data. Our sample was predominantly men due to the inclusion of a VA hospital and it is possible, but unlikely, that women may have differential responsiveness to ACE inhibitors as compared to men. In addition we are unable to examine whether or not there was significant amounts of non-prescription or non-compliance of ACE inhibitor therapy due to economic considerations. Also we are unable to assess factors such as inpatient continuation of the ACE inhibitor or the dose effect due to the design of this study. Further research is needed to examine these factors. Finally, as in any non-experimental study, we are unable to state conclusively that the prior outpatient use of ACE inhibitors is the cause of decreased mortality in this cohort. However, since patients on ACE inhibitors had significantly higher severity of illness scores at presentation we feel that we have good evidence that these medications may have beneficial effects for patients hospitalized with community-acquired pneumonia.
Conclusion
In conclusion, our study finds that prior outpatient use of ACE inhibitors reduces mortality for patients with community-acquired pneumonia. Our results add further strength to the existent recommendations to use ACE inhibitors in patients with congestive heart failure, diabetes mellitus, and renal disease, since these patients are at higher risk for either contracting pneumonia or dying from pneumonia when they do contract it. Future research, especially randomized clinical trials, are needed to examine whether either ACE inhibitors are protective when used in an inpatient setting for patients lacking traditional indications for the use of these medications.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
EMM originated and coordinated the study, obtained funding, contributed to the analysis of the data, and preparation of the paper.
MIR contributed to the design of the study, contributed to the analysis of the data, and preparation of the paper.
AA contributed to the design of the study, contributed to the analysis of the data, and preparation of the paper.
JP contributed to the design of the study, contributed to the analysis of the data, and preparation of the paper.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Dr. Mortensen was supported by Howard Hughes Medical Institute faculty-start up grant 00378-001 and a Department of Veteran Affairs Veterans Integrated Service Network 17 new faculty grant. Dr. Pugh was supported by Department of Veteran Affairs grant HFP98-002. This material is the result of work supported with resources and the use of facilities at the South Texas Veterans Health Care System.
The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
==== Refs
Hoyert DL Arias E Smith BL Deaths: Final Data for 1999 Natl Vital Statistics Report 2001 49 1 113
Gilbert K Fine MJ Assessing prognosis and predicting patient outcomes in community-acquired pneumonia Seminars in Respiratory Infections 1994 9 140 152 7831536
Nakayama K Sekizawa K Sasaki H ACE inhibitor and swallowing reflex Chest 1998 113 1425 9596334
Okaishi K Morimoto S Fukuo K Niinobu T Hata S Onishi T Ogihara T Reduction of risk of pneumonia associated with use of angiotensin I converting enzyme inhibitors in elderly inpatients Am J Hypertens 1999 12 778 783 10480470 10.1016/S0895-7061(99)00035-7
Ohkubo T Chapman N Neal B Woodward M Omae T Chalmers J Effects of an angiotensin-converting enzyme inhibitor-based regimen on pneumonia risk Am J Respir Crit Care Med 2004 169 1041 1045 14990394 10.1164/rccm.200309-1219OC
Gullestad L Aukrust P Ueland T Espevik T Yee G Vagelos R Froland SS Fowler M Effect of high- versus low-dose angiotensin converting enzyme inhibition on cytokine levels in chronic heart failure Journal of the American College of Cardiology 1999 34 2061 2067 10588224 10.1016/S0735-1097(99)00495-7
Alkharfy KM Kellum JA Matzke GR Unintended immunomodulation: part II. Effects of pharmacological agents on cytokine activity Shock 2000 13 346 360 10807009
Skerrett SJ Park DR Anti-inflammatory treatment of acute and chronic pneumonia Seminars in Respiratory Infections 2001 16 76 84 11309715
Moussa K Michie HJ Cree IA McCafferty AC Winter JH Dhillon DP Stephens S Brown RA Phagocyte function and cytokine production in community acquired pneumonia. Thorax 1994 49 107 111 8128397
Puren AJ Feldman C Savage N Becker PJ Smith C Patterns of cytokine expression in community-acquired pneumonia Chest 1995 107 1342 1349 7750329
Bauer TT Monton C Torres A Cabello H Fillela X Maldonado A Nicolas JM Zavala E Comparison of systemic cytokine levels in patients with acute respiratory distress syndrome, severe pneumonia, and controls Thorax 2000 55 46 52 10607801 10.1136/thorax.55.1.46
Meehan TP Fine MJ Krumholz HM Scinto JD Galusha DH Mockalis JT Weber GF Petrillo MK Houck PM Fine JM Quality of care, process, and outcomes in elderly patients with pneumonia JAMA 1997 278 2080 2084 9403422 10.1001/jama.278.23.2080
Niederman MS Mandell LA Anzueto A Bass JB Broughton WA Campbell GD Dean N File T Fine MJ Gross PA Martinez F Marrie TJ Plouffe JF Ramirez J Sarosi GA Torres A Wilson R Yu VL Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention Am J Respir Crit Care Med 2001 163 1730 1754 11401897
Bartlett JG Dowell SF Mandell LA File Jr TM Musher DM Fine MJ Practice guidelines for the management of community-acquired pneumonia in adults. Infectious Diseases Society of America Clin Infect Dis 2000 31 347 382 10987697 10.1086/313954
Fine MJ Auble TE Yealy DM Hanusa BH Weissfeld LA Singer DE Coley CM Marrie TJ Kapoor WN A prediction rule to identify low-risk patients with community-acquired pneumonia N Engl J Med 1997 336 243 250 8995086 10.1056/NEJM199701233360402
Mortensen EM Kapoor WN Chang CC Fine MJ Assessment of mortality after long-term follow-up of patients with community-acquired pneumonia Clin Infect Dis 2003 37 1617 1624 14689342 10.1086/379712
Mortensen EM Coley CM Singer DE Marrie TJ Obrosky DS Kapoor WN Fine MJ Causes of death for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team cohort study Arch Intern Med 2002 162 1059 1064 11996618 10.1001/archinte.162.9.1059
Stone RA Obrosky DS Singer DE Kapoor WN Fine MJ Propensity score adjustment for pretreatment differences between hospitalized and ambulatory patients with community-acquired pneumonia. Pneumonia Patient Outcomes Research Team (PORT) Investigators Med Care 1995 33 AS56 66 7723462
Mortensen EM Restrepo M Anzueto A Pugh J Effects of guideline-concordant antimicrobial therapy on mortality among patients with community-acquired pneumonia Am J Med 2004 117 726 731 15541321 10.1016/j.amjmed.2004.06.028
Klungel OH Martens EP Psaty BM Grobbee DE Sullivan SD Stricker BH Leufkens HG de Boer A Methods to assess intended effects of drug treatment in observational studies are reviewed J Clin Epidemiol 2004 57 1223 1231 15617947 10.1016/j.jclinepi.2004.03.011
|
16159398
|
PMC1236938
|
CC BY
|
2021-01-04 16:30:12
|
no
|
BMC Pulm Med. 2005 Sep 13; 5:12
|
utf-8
|
BMC Pulm Med
| 2,005 |
10.1186/1471-2466-5-12
|
oa_comm
|
==== Front
BMC Plant BiolBMC Plant Biology1471-2229BioMed Central London 1471-2229-5-181613732510.1186/1471-2229-5-18Research ArticleGeneration and analysis of 9792 EST sequences from cold acclimated oat, Avena sativa Bräutigam Marcus [email protected]öf Angelica [email protected] Shakhira [email protected] Gokarna [email protected] Björn [email protected] Olof [email protected] Department of Cell and Molecular Biology, Göteborg University, Box 462, 403 20 Göteborg, Sweden2 Department of Computer Science, Högskolan i Skövde, Box 408, 541 28 Skövde, Sweden2005 1 9 2005 5 18 18 11 3 2005 1 9 2005 Copyright © 2005 Bräutigam 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
Oat is an important crop in North America and northern Europe. In Scandinavia, yields are limited by the fact that oat cannot be used as a winter crop. In order to develop such a crop, more knowledge about mechanisms of cold tolerance in oat is required.
Results
From an oat cDNA library 9792 single-pass EST sequences were obtained. The library was prepared from pooled RNA samples isolated from leaves of four-week old Avena sativa (oat) plants incubated at +4°C for 4, 8, 16 and 32 hours. Exclusion of sequences shorter than 100 bp resulted in 8508 high-quality ESTs with a mean length of 710.7 bp. Clustering and assembly identified a set of 2800 different transcripts denoted the Avena sativa cold induced UniGene set (AsCIUniGene set). Taking advantage of various tools and databases, putative functions were assigned to 1620 (58%) of these genes. Of the remaining 1180 unclassified sequences, 427 appeared to be oat-specific since they lacked any significant sequence similarity (Blast E values > 10-10) to any sequence available in the public databases. Of the 2800 UniGene sequences, 398 displayed significant homology (BlastX E values ≤ 10-10) to genes previously reported to be involved in cold stress related processes. 107 novel oat transcription factors were also identified, out of which 51 were similar to genes previously shown to be cold induced. The CBF transcription factors have a major role in regulating cold acclimation. Four oat CBF sequences were found, belonging to the monocot cluster of DREB family ERF/AP2 domain proteins. Finally in the total EST sequence data (5.3 Mbp) approximately 400 potential SSRs were found, a frequency similar to what has previously been identified in Arabidopsis ESTs.
Conclusion
The AsCIUniGene set will now be used to fabricate an oat biochip, to perform various expression studies with different oat cultivars incubated at varying temperatures, to generate molecular markers and provide tools for various genetic transformation experiments in oat. This will lead to a better understanding of the cellular biology of this important crop and will open up new ways to improve its agronomical properties.
==== Body
Background
Avena sativa (oat) belongs to the Poaceae family. Other cereals in this family are wheat, barley and rye [1]. Wild oats are diploid, but all cultivated oats are hexaploid with an estimated 1C genome size of 13.23 pg, corresponding to about 13000 Mbp [2]. The commercial value of oat is derived both from its high-energy grain and from superior break-crop benefits. Oat plantations also have a comparatively low input demand of insecticides, fungicides and fertiliser due to high disease tolerance and low nourishment requirements of this crop [3]. Today oat is mainly used as animal feed, but it is one of the most promising future cereals in the functional food area. It has unique and well-documented cholesterol lowering effects, as a result of its soluble dietary fibres and high β-glucan content [4-6]. An oat diet greatly improves the well feeling of persons with celiac disease and also reduces the risk of diet-related diseases [7-9]. Oat is rich in natural phenolic antioxidants [10-14], which prevent the development of cardiovascular disease and certain cancers. In Sweden most of the harvested oat is used for animal feed. Only about 35000 tons per year are used for human consumption. However, due to its many health-enhancing properties, the market for oat and processed oat products like oat milk for human consumption is rapidly growing.
Many European countries grow winter oat, i.e. oat that is sown in the autumn and survives the winter in the field. Winter oat therefore has a longer growth season compared to summer varieties, allowing an earlier harvest and giving a higher yield. However, inherently oat is not as winter hardy as rye, wheat and barley. Due to the harsh climate in the Scandinavian countries, winter oat is therefore not grown there. Based on the English experience with winter oat, a Swedish winter oat would probably increase the yield of the harvest by at least 30% (John Valentine, IGER, UK, personal communication). In addition, since oat is the most important rotation crop for wheat and oil crops, an early harvest of winter oat would mean an earlier sowing also of the rotation crops, resulting in increased yields also for these crops. To develop a winter oat suitable for the Scandinavian climate is therefore of high priority (Anders Jonsson, The SwedishFarmers Supply and Crop Marketing Co-operative, personal communication). Since cold hardiness is a quantitative trait controlled by several genes [15], the traditional plant breeding programs have so far been of limited success in improving the cold hardiness for any of the important crop species [16] and the Swedish oat breeders have more or less given up their efforts to produce a Swedish winter oat (Rickard Jonsson, Svalöf Weibull AB, personal communication).
A cost efficient and rapid way to obtain new data from an organism with a large, complex and unknown genome is through partial sequencing of randomly selected cDNA clones. The resulting collection of expressed sequence tags (ESTs) reflects the level and complexity of gene expression in the sampled tissue and will also give an insight into gene structure of the chosen organism. This not only leads to the identification of a number of genes from the new organism that have known or putative functions but also to the discovery of completely novel, previously unknown putative proteins.
In this work, starting from 9792 EST sequences, we identified 2800 putative oat genes, several of which showed similarities to genes previously defined as cold stress related or involved in transcriptional regulation, signal transduction or metabolism. Several sequences that could represent new, unknown and unique oat genes were also identified. This data will now be used to study cold-acclimation in oat, to identify key genes in regulating winter survival, to produce molecular markers to facilitate the breeding for winter hardiness and to construct transgenic oat with increased freezing tolerance. These experiments will increase our general knowledge about the physiology of cold acclimation in plants in general and in oat in particular and in the long term allow the development of a Scandinavian winter oat.
Results
Cold-acclimation
120 individual, four-week old, greenhouse-grown plants of the winter oat varieties Gerald, 83-48 CH and the spring oat varieties Matilda and Birgitta were incubated in the dark at +4°C (± 0.5°C) for 12 and 24 h and thereafter moved to -15°C for 1, 2, 4, 8 or 16 h. Another 120 individual plants were transferred directly at -15°C for 1, 2 4, 8 or 16 h. After the incubation period the plants were transferred back to the greenhouse and allowed to recover for one week. Visible freezing damage was then scored on a scale from 1 to 5, where 1 means no visible damage and 5 means a dead plant (Figure 1A–D). The damage was less severe on the plants that were incubated at +4°C prior to freezing than on plants directly transferred to -15°C (Table 1). Thus, oat plants, as expected, are able to cold-acclimatise to some extent. A clear difference could be seen between summer and winter varieties, however, the latter being more cold adaptable and more winter hardy both before and after acclimation (Table 1).
To confirm cold acclimation on the molecular level a known marker for cold acclimation was investigated. Total RNA was isolated from 3 week-old plants of the winter variety Gerald after incubation at +4°C for 4, 16, 32 and 64 hours. An oat sequence, corresponding to the wheat COR410 gene, known to be cold inducible in wheat [17], was amplified by PCR from genomic oat DNA. A northern analysis using RNA isolated from cold induced plants and oat COR410 as a probe, showed that the COR410 gene expression could be detected after 4 h and then remained strongly expressed even after 64 h of cold incubation (Figure 1E).
EST sequencing and UniGene set construction
Since we aimed for both early and late cold inducible genes in our EST analysis, a cDNA library was constructed from pooled RNA samples isolated from oat plants at different stages of cold acclimation. After plating the library, bacterial colonies were randomly picked and 9792 single-pass sequence reactions performed on cDNAs present in plasmids from these clones. This resulted in 8508 high-quality ESTs of 100 bp or longer, with an average sequence length of 710 bp. The 8508 ESTs were assembled into 1100 contigs and 2616 singletons, giving 3716 candidate genes. All ESTs that contained rRNA, mitochondrial or chloroplast sequences were excluded from further analysis. Finally, the number of redundant sequences in the candidate gene set was eliminated. This was done by a comparative BlastX analysis of the candidate gene set to the NCBI non-redundant (nr) protein database (see "Materials and Methods"). This reduced the candidate gene set to 2800 transcripts with an average sequence length of 800 bp (Figure 2). These final genes were denoted as the AsCIUniGene (Avena sativa cold induced UniGene) set.
Annotation and functional classification
The annotation of the AsCIUniGene set is based on homology. Each gene in the AsCIUniGene set inherited the annotation from the best match found after a BlastX search against nr protein database at NCBI. An expectation value (E value) threshold of 10-10 was used. All sequences, in total 427, that had E values above this threshold were annotated as unknown.
The AsCIUniGene set was functionally classified by identifying every individual sequence in the set by the protein in the Munich Information Centre for Protein Sequences (MIPS) Arabidopsis database (MATDB) that gave the highest BlastX score. Since every gene in MATDB is assigned to at least one of the 28 different classes defined by the MIPS functional classification system, this proved to be a fruitful way to assign putative functions for sequences in the AsCIUniGene set. One drawback with the MIPS classification system is that it is somewhat crude since many of the functional classes were assigned automatically. Therefore, in addition to the automated process, we manually inspected the functional classification assignments for all the proteins and re-assigned them if necessary. In this way all of the 2373 oat sequences, out of which approximately 400 were manually assigned, were finally classified into the previously mentioned MIPS classes. The remaining 427 sequences (with E values > 10-10 in the BLASTX search versus nr) could not be assigned either automatically or manually and where therefore classified as unknown proteins (Figure 3).
The most abundant ESTs
To analyze the most abundant ESTs from the cold induced cDNA library we grouped the sequences by means of the KOG (Clusters of Eukaryotic Orthologous Groups of Proteins) database. This database currently includes 7 eukaryotic genomes, including Arabidopsis. This gave a functional annotation of sequences based on orthologous proteins [18]. In addition we used complementary databases to annotate our sequences like FOGs (Fuzzy Orthologous Group), which contains proteins with promiscuous domains that has not been assigned a KOG identity due to unclear orthologous relationships, TWOGs, which contains provisional clusters of proteins that are represented in two genomes and LSEs, which contains proteins that are lineage-specific expansions of paralogs present in the KOG database. The classification was based on best homology match of BlastX searches against Arabidopsis protein sequences where an expectation value (E value) threshold of 10-10 was used. Proteins annotated in this way were termed "KOG-TWOG-LSE". Since not all Arabidopsis proteins are represented in the KOG database, not all ESTs could be annotated with a KOG-TWOG-LSE. Oat ESTs that had a homology match in Arabidopsis but not a KOG-TWOG-LSE annotation, inherited the annotation from MIPS annotation in MATDB. In addition, several of the sequences did not have an Arabidopsis homolog match with an E value above the threshold. These sequences were annotated with the best homolog match from a BlastX search against the nr protein database at NCBI. Again, an E value threshold of 10-10 was used.
As a comparison, 2189 EST sequences from a non-induced oat leaf library [19] were analysed in the same way. As can be deduced from Table 2, the non-induced and the induced leaf libraries are quite different. In both libraries, chlorophyll a/b binding, ribulose 1, 5-bisphosphate carboxylase/oxygenase small chain and ribulose bisphosphate carboxylase/oxygenase activase (InterPro accession numbers IPR001344 and IPR000894) were the most expressed gene families, although in a different order (Table 2). The ribulose 1, 5-bisphosphate carboxylase/oxygenase sequence was the most common in the non-induced library, perhaps reflecting the higher photosynthetic activity in these plants. Interestingly, several cold related proteins were among the most abundant ESTs in the cold-acclimated library but were not represented among the most abundant ESTs from oat grown under optimal conditions. The cold-induced COR410 (Wcor410) is a dehydrin [17]. Dehydrins are expressed during water-deficit and cold stress. The cold-responsive LEA/RAB-related COR protein (Wrab17) belongs to group-3 of LEA-proteins and has previously been established as induced by cold [20]. Other interesting proteins in this context are the auqaporin PIP and hydroxyproline-rich glycoproteins. The aquaporins are membrane proteins that facilitate water transport across biological membranes. In Arabidopsis there are 13 members of aquaporins that belong to the plasma membrane intrinsic protein (PIP) subgroup. Recently, a study has shown that the PIP proteins are either up- or down-regulated in response to various abiotic stresses [21]. Hydroxyproline-rich glycoproteins have a strong homology to OSR40 proteins in Oryza sativa, which have previously been shown to be up-regulated by salt stress [22]. None of these genes were present in the non-induced library, indicating the great enrichment for cold stress related genes in the induced library. Since only genes that were annotated as cold induced, cold acclimated or as cold regulated were included, other stress related genes that indirectly also are involved in cold stress responses are most likely also enriched in this collection.
Cold-regulated genes
In the functional classification sequences belonging to four categories; "Cell Rescue, Defence and Virulence", "Cellular Communication/Signal Transduction Mechanism", "Metabolism" and "Transcription" were considered to be of great potential interest for cold acclimation. Together these categories were represented by 931 entries in the AsCIUniGene set, corresponding to more than 30% of all genes (Table 3). To increase the resolution of this analysis and to improve the identification of putative cold-regulated genes we built a separate database with proteins previously reported in the literature to be directly involved in cold stress-related processes [23-26]. This "cold stress database" (CSDB) presently comprises 545 entries. In a BlastX search with the AsCIUniGene set against CSDB we identified 398 sequences in the AsCIUniGene set that showed significant homology (E ≤ 10-10) to genes in the CSDB (Table 3). Thus, 14.2% of all the genes in the AsCIUniGene set seem to be cold stress related. Looking just at the two classes that we consider as most important for the cold acclimation process (Cell rescue, defence and virulence; Cellular communication and signal transduction) almost 40% of all sequences were homologous to potential cold stress related proteins. In the "Metabolism" and "Cellular Transport and Transport Mechanisms" classes, cold stress related genes were also overrepresented (approx. 25%).
To analyze whether this high proportion of cold regulated genes in relation to the total number of genes in the AsCIUniGene set was similar to other EST collections derived from cold acclimated plants, we downloaded and analysed EST datasets from cold acclimated wheat and barley. The datasets were clustered and assembled with the TGI clustering tool (see Material and Methods). This resulted in a TaCIUniGene set of 2894 genes and a HvCIUniGene set of 3932 genes. In addition, as a comparison the same oat ESTs collection derived from leaves of plants grown under optimal conditions (Table 2) was clustered and assembled into a AsNIUniGene set of 1445 sequences. These three UniGene sets were then searched against the CSDB. This showed that the proportions of cold stress related genes were 9.6% in the TaCI candidate gene set and 11.1% in the HvCI, but only 5.1% in the AsNI UniGene set (Table 4). Thus the AsCI UniGene set is quite different from previous oat EST collections and also the EST collection that contains the highest proportion of cold induced genes of all cereals.
Phylogenetic analysis of AP2 containing proteins
Among the 190 sequences in the transcription class (Table 3), 107 were found to be homologous to transcription factors. Remarkably, 51 of these 107 sequences were homologous to genes in the CSDB, representing almost 48% of all transcription factors found in the AsCIUniGene set. To investigate this a bit more, the transcription factors were further classified into 14 different transcription factor families (Table 5). Dominating among these were genes encoding AP2 domain, homeodomain and zink finger proteins. Proteins belonging to the CBF/DREB1 transcription factor family have previously been shown to be the regulators of the majority of cold-response genes. The CBF/DREB1 family belongs to the AP2/ERF super family [27] and in Arabidopsis the AP2/ERF super family comprises 145 proteins. Based on similarities in their DNA-binding domains, these proteins have been subdivided into the AP2, RAV, DREB and ERF subfamilies and one family with the remaining proteins. In the AsCIUniGene set, we found 11 sequences belonging to the AP2/ERF superfamily (Table 5). The AP2/ERF domain of these 11 Avena proteins were more closely analysed and also compared to 45 previously described AP2/ERF containing proteins [28]. The resulting phylogenetic tree revealed that 4 of the oat AP2/ERF proteins belonged to the DREB subfamily, 2 to the AP2/RAV subfamily and the remaining 5 sequences to the ERF subfamily (see Figure 4 and Table 6). From the analysis it can also be deduced that the oat CBF/DREB1 proteins are most closely related to the monocot CBF/DREB1 proteins (Figure 4).
To further analyse relations between oat and other monocot CBF/DREB1 proteins, a multiple alignment of AP2/ERF domains from AsCBF1, AsCBF2, AsCBF3, AsCBF4, OsDREB1A, OsDREB1B, HvCBF1, HvCBF2, HvCBF3, ScCBF and TaDREB was made. In addition, the Arabidopsis CBF/DREB sequences AtCBF1, AtCBF2 and AtCBF3 were included to further elucidate the relation between dicots and monocots in this respect (Figure 5). In previous studies it has been reported that in particular two amino acids, a valine at position 19 (V19) and a glutamic acid at position 24 (E24) (Figure 5) in the AP2/ERF domains of Arabidopsis have important roles in determining DNA-binding specificity [27]. The AsCBF proteins have the conserved valine in the V19 position but not the glutamic acid in the E24 position (Figure 5). Instead, a valine is conserved in this position. This feature is in fact shared among all included monocot CBF/DREB proteins (Figure 5). The monocot CBF/DREB1 proteins could be further divided into three subgroups (Figure 4). The first subgroup (G1) contained the AsCBF3, HvCBF1, HvCBF2, ScCBF, and TaDREB proteins, the second subgroup (G2) contained the AsCBF1, AsCBF2 and HvCBF3 proteins and the third subgroup the remaining AsCBF4, OsDREB1 and OSDREB2 proteins (Figure 4). This grouping is based on differences in aa at positions 10, 18 and 39 between the different proteins (Figure 4, Figure 5). At position 10 the G1 proteins have a basic arginine (R) residue, the G2 proteins a hydrophilic asparagine (N) and the G3 proteins N, hydrophilic serine (S) or glycine (G) residues (Figure 5). In position 18 there is a basic residue, an arginine (R) in G2 and G3 proteins whereas G1 proteins have a hydrophilic residue, a glutamine (Q) (Figure 5). Finally G2 proteins have a hydrophilic tyrosine (Y) in position 39 while in G1 and G3 proteins this position is occupied by a hydrophobic phenylalanine (F) except for AsCBF4, which has a basic histidine (H) (Figure 5).
Expression of the AsCBF genes
To explore whether the four identified oat AsCBF genes were cold induced with similar kinetics as other previously described CBF genes an expression analysis was performed. Multiplex RT-PCR was run on total RNA isolated from leaves of three weeks old plants cold induced (to +4°C) at time points between 15 min to 24 h using gene specific primers for AsCBF1, AsCBF2, AsCBF3 and AsCBF4 respectively. Total RNA isolated from untreated plants at the same time points were used as a comparison. An oat actin gene (AsActin) was also amplified from the same RNA samples as a loading and RNA quality control. To define conditions where the PCR amplification was in the exponential phase, several experiments with 30, 35 or 40 cycles were performed. This showed that the different AsCBF genes all were induced by the cold treatment but that they were differentially regulated (Figure 6). The AsCBF1 gene was not detectable at time 0, was induced after about 30 min, peaked at 4 hours and was completely shut off after 24 h. The AsCBF3 gene had quite a different expression pattern. It was weakly expressed also in non-induced plants, but was in addition rapidly induced already at the first time point after 15 min. The expression levels continued thereafter to increase, peaked after about 4 hours but still showed an elevated expression at 24 h. The AsCBF4 gene was slowly induced and not detected at all until after 4 hours. Unlike the others its expression peaked a bit later, after 8 hours and then had completely declined after 24 h. Despite several attempts, using different primer pairs we could not obtain a reproducible expression pattern of the AsCBF2 gene (data not shown). Thus, the different oat AsCBF genes seem to be active during different phases of the cold acclimation process and therefore perhaps induce different downstream gene programs. We are presently addressing this issue more specifically.
Identification of microsatellites
Using the Sputnik program and threshold values as specified in the Methods section, we searched for potential microsatellite (SSR) sequences in the 3716-candidate gene set. In total, 399 di- to pentanucleotide SSRs that fulfilled the criteria of the search were identified. This corresponds to approximately one microsatellite per 13 kb of sequence. Using the same methods and thresholds, Cardle et al. (2000) found on average one SSR per 14 kb of EST data in Arabidopsis. This indicates that the SSR frequency in oat is similar to that in Arabidopsis. In the oat sequence collection, tri-nucleotide repeats were the most commonly found followed by di-nucleotide repeats (Figure 7), which again matches the results from Arabidopsis [29]. Except for two exceptional TA/AT SSRs of length 45 and 55 bp, SSR lengths ranged from 15 to 25 bp, with 16 and 17 bp being the most common.
Work is now in progress to determine which of these SSRs can be reproducibly amplified by PCR, are polymorphic, and can be linked to a phenotypic marker. The vast majority of the oat SSRs were found in non-coding DNA. Since they nevertheless represent actively transcribed genes, we expect that several of these will turn out to be useful markers for breeding.
Discussion
Plant expressed sequence tags (ESTs) have proven to be valuable tools in molecular biology research and a number of collections from many different plant species are now publicly available [19]. In cereals, which are the most important food providers on earth, several major EST sequencing projects have been carried out. At the time of writing, there are 284 779 publicly available ESTs from Oryza sativa (rice), 562 786 from Triticum aestivum (wheat) and 367 768 from Hordeum vulgare (barley). In contrast, there are only 7 624 entries for Avena sativa (oat) and no sequences from cold acclimated oat are available. Obviously, there is a great need for more EST sequencing also on this important crop. Here we contribute an additional 9 792 sequences, originating from cold-acclimated oat, to the research community. Since we were mainly interested in genes involved in the perception, signal transduction and early regulation of cold acclimation, we focused on short incubation times from a few minutes to 24 h. Already after 12 hours acclimation, there was a clear difference in freezing tolerance between acclimated and non-acclimated plants and winter varieties were more tolerant than spring varieties (Table 1). To confirm that cold induced genes were overrepresented in these plants, a northern analysis was performed on an oat gene corresponding to the previously described cold induced wheat COR410 gene on RNA isolated from several different time points during cold acclimation at +4°C. This revealed that the diagnostic COR410 gene was cold induced also in oat and, interestingly, the peak expression level was higher in the winter varieties (Figure 1 and Table 1). The same tendency with earlier induction and higher expression levels was also seen for other cold induced genes (data not shown).
Pooled leaves from confirmed cold induced plants were used for cDNA construction and EST sequence generation. Since leaves were used as the RNA source, the most common ESTs in our collection represent various genes involved in photosynthesis, like chlorophyll a/b binding protein, ribulose 1, 5-bisphosphate carboxylase/oxygenase, ribulose bisphosphate carboxylase/oxygenase activase, photosystem I reaction centre protein, fructose-bisphosphate aldolase, carbonic anhydrase/carbonate dehydrase and photosystem II oxygen-evolving complex proteins. Other well-represented sequences are those encoding ribosomal proteins (Table 2). However, among the 20 most expressed gene families, dehydrin was also present, indicating that our collection indeed represents plants with a cold stressed induced condition. This was confirmed by a direct comparison to an EST set derived from leaves of non-induced plant. In this collection, dehydrin and other cold induced genes were not among the 20 most highly expressed.
From our cold induced EST collection, an AsCI UniGene set of 2 800 genes was identified. Of these, 1 726 could be placed into the functional groups defined by MIPS (Figure 3), leaving a relatively large proportion of the genes (approx. 40%) outside of this classification. Perception of the stress stimuli, transduction of the stress signal and a molecular response are necessary activities if the plant is to react to abiotic stress. In oat, however, very little is known about cold stress response at this level, although the general mechanisms are probably similar in all plants. In order to better identify oat genes involved in the cold response we therefore created a database denoted CSDB (cold stress data base), in which all genes available from the public domain and classified as cold stress responsive or cold induced were collected. When the sequences in the CSDB were compared to the oat AsCIUniGene set we found that 398 sequences matched, indicating that at least 14% of all the genes in the AsCIUniGene set are involved in cold stress. Among these, sequences encoding activities related to perception, signal transduction and transcriptional regulation were overrepresented (Table 3). From the oat EST collection generated from leaves of three weeks old oat plants grown under green house conditions we created a UniGene set of 1445 different transcripts using the same tools as with the AsCIUniGene set. This non-induced leaf set was denoted AsNIUniGene. When the CSDB was searched with AsNIUniGene only 5.1% of the genes were found to be similar (Table 4), a dramatic difference to the AsCIUniGene set. These studies were extended to EST collections from cold acclimated wheat and barley. By creating UniGene sets (TaCIUniGene and HvCIUniGene) both these collections were analysed in the CSDB. We then found that the amount of cold stress related genes were around 10% in both the wheat and barley collections (Table 4). Generalising, it seems like that at least 10% of all expressed genes in cold acclimating plants are devoted to various cellular responses needed to prepare the plant to freezing temperatures. The cold induced oat gene collection will now be a valuable new asset in further analysis of such genes.
Our functional analysis of the AsCIUniGene set showed that transcription factor genes were represented by 107 sequences, belonging to at least 14 different families (Table 5). Of these, 51 were homologous to cold-induced genes from other systems. Of special importance for cold acclimation is the CBF transcription factor family. Genes in this family regulate several different downstream genes, including the COR genes [16,30]. However, this regulation is complex and several different CBF genes are involved. From the AsCIUniGene set we identified four oat CBF genes, denoted AsCBF1, AsCBF2, AsCBF3 and AsCBF4. Our phylogenetic and multiple alignment analysis showed that all four belonged to the monocot DREB subfamily of ERF/AP2 domain proteins. The AsCBF1 and AsCBF2 genes were very closely related, while the AsCBF3 and AsCBF4 genes were somewhat more distantly related to each other and also belonged to a different clade than the AsCBF1 and AsCBF2 genes (Figure 4 and 5).
To investigate the expression profile of the AsCBF genes, we designed gene specific primers and by RT-PCR analysis showed that these genes indeed were cold induced, but that their expression patterns were different. Their expression ranged from early induction already after 15 min (AsCBF3) to induction after 1 h (AsCBF4) and from peaking at 4 h (AsCBF1 and AsCBF3) to peaking at 8 h (AsCBF4) (Figure 6). The AsCBF3 was particularly interesting since it was weakly constitutively expressed, showed a clear increase in expression after cold treatment and still expressed after 24 h. Despite several attempts using different primer pairs we could not obtain a reproducible expression pattern of the AsCBF2 gene. The reason for this is presently not known. The complex regulation of the AsCBF genes is different from what was previously described in Arabidopsis [31] where the AtCBF1, AtCBF2, and AtCBF3 genes follow more or less the same expression pattern with a rapid induction after 15 min and a peak after 2 h. This indicates that CBF factors have intricate and different individual roles in inducing and maintaining cold acclimation in oat. This is corroborated by preliminary data from barley. This cereal has at least 10 different genes encoding CBF factors, which are all differentially regulated (Eric Stockinger, Ohio State University, personal communication). Thus, a more detailed analysis of the structure and regulation of CBF genes in cereals may reveal new pathways of cold induction, not present in Arabidopsis.
A number of genes with hitherto unknown functions were identified in the AsCIUniGene set. These were divided into two groups, one in which homologous or similar genes from other systems exist and one where no significant similarities could be found to any other sequence, i.e. genes that could be oat specific. In order to rule out that small "non-real" peptides contributed to this group, only sequences with open reading frames of 100 aa or more were included. Work is now in progress to elucidate which of the 427 oat specific unknown genes that are induced by cold stress, drought stress or combinations of different stress factors. Assuming that approx. 10% of these sequences are cold related, more than 40 completely new oat genes involved in cold acclimation will be present in this collection. Such genes are potentially very interesting and could encode hitherto uncharacterised proteins or regulatory factors involved in cold-adaptation and freezing protection
Microsatellites (SSRs) are excellent DNA markers for genetic mapping, since they are polymorphic, abundant, show a co-dominant inheritance and are easy to analyse by PCR [32]. SSRs have therefore been widely utilized in plant genomic studies [33-36]. They are especially advantageous when there is a need to track desirable traits in large-scale breeding programs and when defining anchor points for map-based gene cloning strategies. However, only a few oat SSRs are currently available. Here we identify approximately 400 potential oat SSRs, the majority present in the non-coding part of the EST sequence. Work is now in progress to optimise primers for these SSRs and to identify the ones that give reliable PCR products and are polymorphic. Crude genetic maps have been developed for both diploid [37,38] and hexaploid oat [39], but these maps need to be improved [40]. The best SSR markers will therefore be mapped to the oat genome, and linked to valuable genetic markers.
The AsCIUniGene set will now be used to fabricate an oat biochip carrying all 2800 identified genes. In addition, by constructing subtractive libraries, more cold-related ESTs will be generated. Various expression studies will be performed and genes from our collection that show a rapid induction to either cold or drought will be selected for further analysis. We are especially interested in those genes in our EST collection that show a very rapid induction at +4°C and have DNA binding properties. Especially promising genes will be tested in transgenic Arabidopsis and oat systems [41,42] and by complementing chosen Arabidopsis T-DNA knock-out mutants.
Conclusion
A UniGene set of 2800 genes was produced from a cold induced oat cDNA library. Further analysis revealed that genes related to cold stress were overrepresented in this library and that several genes could encode hitherto unknown functions. RT-PCR analyses of CBF transcription factor genes revealed that they are differentially expressed in oat and therefore might regulate different cold pathways. Approx. 400 potential SSR markers are also present in the collection, several in non-coding regions and in close vicinity to genes involved in regulating cold acclimation.
Methods
Plant growth
Oat plants, Avena sativa v. Gerald, 83-48 CH, SW Matilda and SW Birgitta were obtained from the SW-collection (Svalöf Weibull AB, Landskrona, Sweden). Gerald and 83-48 are English winter varieties while Matilda and Birgitta are Swedish spring varieties. Seeds were germinated in 2-liter pots filled with fertilized and pressed peat. Plants were cultured in a greenhouse under natural light supplemented with metal halogen lamps, giving a photon flux density of 240 μmol per m2 per sec. The photoperiod was 18 h, the day/night temperature was 20/12°C and the relative humidity about 70%. The plants were watered as needed.
Cold induction experiments
To investigate the cold acclimation capacity of the chosen oat varieties, 24 pots with 10 seeds each of Gerald, 83-48 CH, Matilda, and Birgitta were prepared. About three weeks after germination, when each plant had produced 3 – 4 leaves, pots were moved to a dark cold room at +4°C (± 0.5°C) and incubated for 12 and 24 hours. After this period the pots were moved to -15°C (± 1°C) for 3 h, 6 h and 12 h. In addition, plants were moved from the greenhouse directly to -15°C, and incubated for 3 h, 6 h and 12 h. After the cold incubation period, the plants were moved back to the greenhouse for recovery. One week later the cold damage was visually scored.
Total RNA preparation
Winter oat (Gerald) was germinated and grown for three weeks in the greenhouse. They were then incubated in the dark at +4°C (± 0.5°C) for 4, 16, 32 or 48 hours. At every timepoint, leaves were randomly picked from several individual plants and pooled. RNA was extracted from pooled leaves essentially as described by Chang et al. (1993). Tissue isolates were ground in liquid nitrogen, transferred to 65°C CTAB extraction buffer (2% CTAB [hexadecyltrimethylammonium bromide] [Sigma], 2% PVP [polyvinyl pyrrolidone, intrinsic viscosity 29–32] [Sigma], 100 mM Tris-HCl [pH 8.0], 25 mM EDTA, 2.0 M NaCl, 0.5 g per L spermidine, 2% β-mercaptoethanol), and extracted twice in equal volumes of chisam (phenol:chloroform:iso-amyl-alcohol 1:1:24). RNA was precipitated overnight at 4°C by adding 0.25 v/v 10 M LiCl. The precipitate was dissolved in 1 × SSTE (1.0 M NaCl, 0.5% SDS, 10 mM Tris-HCl, 1 mM EDTA), pH 8.0, extracted with an equal volume of chisam, precipitated with two volumes 99.5% ethanol, and re-suspended in distilled water treated with DEPC. Total RNA of each sample was quantified spectrophotometrically at OD260. An OD260 of 1 corresponded to 40 μg/ml RNA. Subsequently, the RNA was precipitated and re-suspended in DEPC-treated distilled water to a final concentration of 1 mg/ml.
Northern hybridization
Ten μg of total RNA were denatured with glyoxal/DMSO [43] and separated on a 1% agarose gel. The RNA was blotted onto a nylon membrane (Boeringer-Mannheim) and hybridized in Church hybridization buffer [44]. An oat sequence, similar to the wheat COR410 gene was used as a probe. This was isolated by PCR amplification from oat genomic DNA using the forward primer 5'-ATGGAGGATGAGAGGAGCAC-3' and the reverse primer 5'-TTTCTTCTCCTCCTCGGGC-3'. Primer design was based on the wheat sequence. Amplification resulted in an 530 bp sequence which was verified by DNA sequencing (data not shown). The fragment was labelled with 32P-dCTP (Amersham), using a random hexanucleotide mix and labelling-grade Kleenow enzyme (Boeringer-Mannheim). Stringency washes were performed at 65°C for 2 × 5 min in 2 × SSC, 0.5% SDS and for 4 × 5 min in 0.2 × SSC, 0.1% SDS. Membranes were exposed to X-ray film (Du Pont Medical Scandinavia AB).
cDNA library construction and EST sequencing
Total RNAs isolated from plants incubated at +4°C for 6, 12 and 24 hours were pooled. The RNA pooled preparation was sent to MWG Biotech (Germany) where cDNA libraries were constructed, cDNA cloned into the pSPORT1 vector [45] and EST sequencing was performed.
Bioinformatic tools
All similarity searches were batch executed locally using the BlastN, BlastX or TBlastX tools [46], all included in the BLASTALL program package [47]. Transeq, a program from the EMBOSS package [48,49] was used to translate DNA sequences into protein sequences. Conserved domain search (CD-search) was performed against the Conserved domain database (CDD) at NCBI [47] using the Reversed position specific Blast (RPS-BLAST) algorithm with translated ESTs. InterProScan [50,51] was used to scan translated ESTs for protein signatures in the InterPro member databases. For the multiple alignments we used ClustalW [52], included in the MacVector 7.2.2 package (Accelrys Inc). The phylogenetic tree was constructed by means of the MacVector 7.2.2 tool kit using the neighbour-joining (NJ) algorithm. Appropriate PERL scripts were written in order to pipeline the process of running tools in sequence, parsing result files and loading the results into the database. All data and results are stored in a PostgreSQL database.
Data sets and treatment
In this paper started off with four different primary sequence data sets. The first set was the 9792 ESTs from cold acclimated oat, which was denoted the Avena sativa Cold Induced (AsCI) data set. The second data set comprises 2189 ESTs [53], which originated from untreated green leafs of 3 weeks old oat plants and was denoted the Avena sativa NonInduced (AsNI) data set. The third data set, which contains 4337 sequences originates from cold acclimated wheat [53] was denoted the Triticum aestivum Cold Induced (TaCI) data set and the final data set comprises 5418 sequences from cold stressed barley plants [53] and was denoted the Hordeum vulgare Cold Induced (HvCI) data set.
EST clustering and assembly
The AsCI data set was filtered, clustered and assembled with the Paracel Transcript Assembler™ (PTA) program (Paracel, Pasadena, CA), which integrates quality filtering, clustering, and assembly into a single pipeline. The filtering step includes masking of vector sequence, low-complexity, low-quality, repeats and poly(A) regions. In the next step clustering was performed. Here PTA utilizes the Haste algorithm in an all-versus-all sequence comparison. The criterion set for clustering sequences together was an alignment of a minimum of 100 bases and with at least 93% similarity between the aligned sequences. Sequences that did not fit into such clusters were defined as singlets. In the assembly step PTA uses CAP4, which is a refinement of the CAP3 algorithm [54]. Sequences that did not fit into a contig were also defined as singlets. Finally, ESTs in singlets that had passed the filters but had an unmasked sequence < 100 bases were discarded. The resulting singlets and contigs represented the AsCI candidate gene set.
The other data sets where clustered and assembled into candidate gene sets using the TGI clustering tool [55]. The clustering was performed by a slightly modified version of NCBI's MegaBlast program [56] and the resulting clusters were assembled using CAP3.
Most abundant ESTs
Individual ESTs were first annotated by the best BlastX homolog match against Arabidopsis thaliana, where an E value of < 10-10 was used. The A. thaliana proteins were retrieved from the MIPS Arabidopsis database (MATDB). Thereafter the annotation given in the KOG database [18] was retrieved for each A. thaliana protein. For those A. thaliana homologs that did not have a KOG annotation, the EST sequence inherited the annotation from MATDB. Those sequences that did not have an A. thaliana homolog above the threshold were annotated with the best homolog match from a BlastX search versus the nr database at NCBI. Again an E value of < 10-10 was used.
UniGene set determination
Non-redundant sets of genomic singlets and contigs (UniGene sets) were created in a two-step procedure. First sequence information derived from rRNA, chloroplastic DNA or mitochondrial DNA was identified by comparison to homologous Arabidopsis sequences (accession nr. X52322, AP000423, and Y08501/Y08502 respectively) using BlastN. In this way, sequences containing rRNA or mitochondrial DNA were separated from the genomic sequences.
The second step was a BlastX search of the non-redundant (nr) protein database. The accession numbers and E values of the best matches were extracted from the result file. The criterion used for a sequence to be identified as non-redundant was based on a unique best match based on the accession numbers from the BlastX search. If two or more query sequences resulted in best matches with identical accession numbers they were sorted according to their E values. Only the sequence with the lowest E value was included in the UniGene set.
Annotation and functional classification
The UniGene sets were annotated based on the results of BlastX searches of the nr database. The definition line of the Blast match was used as a description of the putative function of the UniGene gene. An E value threshold of 10-10 was used and UniGene genes that did not meet this requirement were annotated as unknown.
Our functional classification of individual genes followed the functional categories as defined in the Munich Information Centre for Protein Sequences (MIPS) Arabidopsis thaliana functional catalogue (MATDB; downloaded from ). To create a semi-automated functional classification pipeline a two-step procedure was developed. First, a BlastX search was performed with the UniGene set versus the MATDB, requiring an E value of < 10-10. Locus name and E value of the best match for each gene were extracted from the result file. Secondly, the functional classification was identified by a search with the locus name in the Arabidopsis functional catalogue. Genes that did not meet the criteria for being functionally classified based on the semi-automated procedure were classified manually based on the annotation and the result from a conserved domain search versus the conserved domain database (CDD) downloaded from the NCBI web site.
Identification of microsatellite sequences
A set of 3716 sequences resulting from the clustering and assembly steps, comprising a total of 5.3 Mb of sequence, was searched for microsatellites (simple sequence repeats; SSRs) in the form of mononucleotide repeats of > 15 bp, dinucleotide repeats of > 14 bp, trinucleotide repeats of > 15 bp, tetranucleotide repeats of > 16 bp, and pentanucleotide repeats of > 20 bp, as previously described by [29]. To better locate di- to pentanucleotide repeats, we also used the program Sputnik, developed at the Washington University [57]. This program allows minor imperfections on the SSRs by implementing a scoring system for insertions, mismatches and deletions. To locate mononucleotide repeats we used a simple PERL script developed by ourselves.
RT-PCR
Reverse Transcriptase Polymerase Chain Reactions (RT-PCR) were performed on total RNA prepared from leaves of three weeks old oat plants (variety Gerald), incubated for 0 min, 15 min, 30 min, 45 min, 1 h, 2 h, 8 h and 24 h at +4°C, using the SuperScript™ III One-Step RT-PCR system (Invitrogene™). RNA samples were first DNase treated using the Dnase I Amplification Garde from Invitrogene™. To amplify the different AsCBF genes the following primers were used:
AsCBF1 forward primer 5'-CCACAGTCCACCGTATCAGCAAG-3'
AsCBF1 reverse primer 5'-CGTCTCCTTGAACTTGGTGCG-3'
AsCBF3 forward primer 5'-CGGGCAAAGTTGAGGCAGGC-3'
AsCBF3 reverse primer 5'-TAGGCTCTGGCTCGGCACCTTC-3'
AsCBF4 forward primer 5'-CCCAGCCTTCAGCAGCGTC-3'
AsCBF4 reverse primer 5'-TCTCCACAGTCTCCTCCGTGC-3'
For the AsCBF1 gene a product size of 174 bp was expected, for AsCBF3 104 bp and for AsCBF4 172 bp. The AsActin gene used as a control and was amplified using the forward primer 5'-GCGACAATGGAACTGGC-3 and the reverse primer 5'-GTGGTGAAGGAGTAACCTCTCTCG-3'. In this case the expected product size was 580 bp. The RT-PCR reactions were run according to the manufacturer's instructions and 100 ng total RNA were used in each reaction. A 30-min reverse transcription at 55°C followed by a PCR amplification step with 30, 35 or 40 cycles were used. To verify the outcome of the RT-PCR reactions, equal amounts (30%) of the corresponding RT-PCR reaction mixes were applied on 1% agarose gels containing ethidium bromide (0.5 ng per ml).
Authors' contributions
MB contributed with EST data analysis, phylogenetic analysis, CBF expression experiments and with writing of the paper. AL did the data analysis on the most expressed genes. GC grew oat plants and performed the freezing experiment. SZ did cold induction of oat, RNA preparation, and quality and induction control of the RNA used for cDNA preparation. BO did EST data analysis, SSR identification and writing. OO contributed with planning, supervising and financing of the work and with writing of the paper.
Acknowledgements
This work was supported by grants from the VL-foundation, The SwedishFarmers Supply and Crop Marketing Co-operative, the Swedish Research School in Genomics and Bioinformatics and the Swedish Research Council (VR).
Figures and Tables
Figure 1 Cold acclimation in oat. (A) Four-week old Matilda (1) and Birgitta (2) (spring oat), Gerald (3) and 83-48-CH (4) (winter oat) grown in green house. (B) Plants recovered for one week in green house after acclimation (24 h at +4°C) and incubation at -15°C for 3 h. (C) Plants recovered for one week in the green house after acclimation (24 h at +4°C) and incubation -15°C for 6 h. (D) Plants recovered for one week in green house after acclimation (24 h at +4°C) and incubation at -15°C for 12 h. (E) In a separate experiment four-week old Gerald plants were incubated at +4°C for 0, 4 h, 16 h, and 64 h (indicated below lanes), total RNA was isolated and Northern hybridisations was performed using the oat Cor410 genes as a probe.
Figure 2 Sequence length distributions. Open bars, all 8508 sequenced ESTs; filled bars, the 2800 UniGene sequences. Average lengths were 710,7 and 800,1 bp, respectively. The longest sequences were 1017 and 2639 bp, respectively.
Figure 3 Functional classification of the AsCIUniGene Set. Classification was done both manually and according to the MATDB classification scheme (see Methods). The functional category is indicated by the text associated to respective piece in the diagram. The size of each piece is proportional to the relative abundance to the proteins assigned to this group.
Figure 4 Phylogenetic analysis of ERF/AP2 domains. The tree was built by means of the Neighbour-Joining method using 56 different ERF/AP2 domains. Numbers along the branches correspond to bootstrap values after 1000 replicates. Branches without values have bootstrap values < 55. AC-numbers and source of the sequences are given in Table 6.
Figure 5 ClustalW analysis of AP2 DNA-binding domains. AP2domain of AsCBF1, AsCBF2, AsCBF3 and AsCBF4 were aligned to AP2 domains from CBF related proteins from other plants as indicated in the text to the left. Arabidopsis thaliana AtCBF1, AtCBF2, AtCBF3; Horedeum vulgare HvCBF1, HvCBF2, HvCBF3; Triticum aestivum TaDREB; Oryza sativa OsDREB1A, OsDREB1B; Secale cereale ScCBF (AC-numbers of the sequences are given in table 6). The amino acid residues are coloured based on the functionality, and the colouring schema is as follows: Acidic residues (DE) are red, basic residues (RHK) are blue, hydrophobic residues (AILMFPV) are white and hydrophilic residues (NCQGSTY) are orange.
Figure 6 RT-PCR analysis of AsCBF gene expression in three-week old oat. Plants were incubated at +4°C for the times indicated below the lanes (m, minutes; h, hours). Total RNA was isolated from leaves and PCR reactions were run in 30 cycles using AsCBF1, AsCBF3, or AsCBF4 specific primers, as indicated to the left of the picture. Equal loading and RNA quality was controlled by amplification of the oat AsActin1 gene. A 1 kb ladder was used as a size marker (not shown).
Figure 7 Microsatellite distribution. SSRs were sorted according to repeat motif length. The height of the bar indicates the number of SSRs that were found for each class. Numbers above bars denote average total SSR length.
Table 1 Cold acclimation in oat. Frost damage in non-acclimated (20°C) and acclimated (12 or 24 h at +4°C) oat spring varieties Birgitta and Matilda and winter varieties Gerald and 83-48-CH. Plants were incubated at -15°C for 3 h, 6 h and 12 h and then recovered for 1 week in the greenhouse. Frost damage was recorded on a scale from 1 to 5, where 1 represents no damage and 5 represents dead plants.
-15°C incubation
No acclimation 3 h 6 h 12 h
Birgitta 2 5 5
Matilda 3 5 5
Gerald 1 3 5
83-48-CH 1 5 5
12 h acclimation 3 h 6 h 12 h
Birgitta 3 4 5
Matilda 3 4 5
Gerald 1 2 5
83-48-CH 1 2 5
24 h acclimation 3 h 6 h 12 h
Birgitta 3 3 5
Matilda 3 4 5
Gerald 1 2 5
83-48-CH 1 2 5
Table 2 The 20 most frequent, randomly picked, EST sequences in two different oat leaf libraries. Gene family, EST sequences identified as belonging to the indicated gene family; CI, total number of genes found in the indicated family in the cold induced leaf library; % of total, relative amount of genes in the indicated gene family, NI, total number of genes found in the indicated family in the non-induced leaf library.
Gene family CI % of total NI % of total
Chlorophyll a/b binding 952 11.3 39 1.8
Ribulose 1, 5-bisphosphate carboxylase/oxygenase small chain 657 7.8 210 9.6
Ribulose bisphosphate carboxylase/oxygenase activase 232 2.7 36 1.6
Carbonic anhydrase 2 149 1.8 29 1.3
Fructose-bisphosphate aldolase 144 1.7 2 0.1
Glyceraldehyde-3-phosphate dehydrogenase 107 1.3 6 0.3
Cold-induced COR410 (Wcor410) 82 1.0 0 0.0
Photosystem II oxygen-evolving complex (PsbP1) 58 0.7 11 0.5
LEA/RAB-related COR protein (Wrab17) 58 0.7 0 0.0
Photosystem II oxygen-evolving complex (PsbO2) 52 0.6 4 0.2
Ferredoxin 51 0.6 5 0.2
Photosystem II (PsbR) 50 0.6 4 0.2
Photosystem I reaction centre subunit XI 46 0.5 0 0.0
Glycolate oxidase 39 0.5 0 0.0
Alanine aminotransferase 39 0.5 3 0.1
Phosphoribulokinase 39 0.5 15 0.7
Aquaporin PIP 37 0.4 2 0.1
Glutamine synthetase 36 0.4 8 0.4
Hydroxyproline-rich glycoprotein 34 0.4 1 0.0
Photosystem I reaction centre subunit psaN 33 0.4 8 0.4
Table 3 Cold stress related oat genes. Distribution of genes in the AsCIUniGene set into different functional categories. Functional class, determined as described in Methods; AsCIUniGene set, the set of 2800 different oat genes; CSDB, set of cold related genes extracted from the public domain; % CSR, relative number of oat genes in each family that is similar to cold related genes.
Functional class AsCIUniGene set CSDB % CSR
Cell cycle, DNA processing, cell fate and development 60 19 31.7
Cell rescue, defence and virulence 130 43 33.1
Cellular communication and signal transduction mechanisms 180 80 44.4
Cellular transport and transport mechanisms 138 28 20.3
Cellular organisation 31 5 16.1
Energy 215 5 2.3
Metabolism 431 87 20.2
Protein fate 144 13 9.0
Protein synthesis 93 1 1.1
Transcription 190 57 30.0
Unclassified protein 752 57 7.6
Unknown protein 427 2 0.5
Other 9 1 11.1
Summary 2800 398 14.2
Table 4 Percentage of cold stress related sequences in different UniGene sets. UniGene specifies the different UniGene sets, which were: AsCI, Avena sativa cold induced; AsNI, Avena sativa non-induced; TaCI, Triticum aestivum cold induced; HvCI Hordeum vulgare cold induced. Amount, refers to the number of sequences that were analysed. In CSDB, indicates how many of the total genes that also were present in the collection of cold stress related genes. Cold related (%), gives the percentage of genes in each set that were cold stress related.
UniGene Amount In CSDB Cold related (%)
AsCI 2800 398 14.2
AsNI 1445 74 5.1
TaCI 2894 277 9.6
HvCI 3932 437 11.1
Table 5 Classification of oat transcription factors. Distribution of the 107 transcription factors, identified in the AsCIUniGene set, in different families classified according to Reichmann et al, 2000.
Family No of genes
AP2/EREBP 11
bHLH 2
bZIP 9
CONSTANS B-box 5
DOF 3
Homeodomain 12
IAA 6
Leuzin zipper 3
MADS box 3
MYB 9
SCARECROW 3
WRKY 4
Zink finger 23
Other transcription factors 14
Total 107
Table 6 List of sequences containing the ERF/AP2 DNA-binding domain. The list of ERF/AP2 sequences used in the phylogenetic analysis (figure 4), some of the sequences were also used in the ClustalW alignment (figure 5). The first two letters in the protein name represent the initial letters of the Latin binomial, followed by the gene abbreviation. Each sequence is assigned to one of three different subgroups of the ERF/AP2 superfamily. A. sativa sequences are grouped according to our phylogenetic analysis (Figure 4). All GI accession numbers correspond to protein sequences in gene bank at NCBI [53], and the EMB accession numbers correspond to EST sequences available in the EMBL-nucleotide sequence database [58].
Species Gene name Group Accessions Reading farme1
Arabidopsis thaliana AtABI4 DREB gi|15225661
AtAPETALA2 AP2/RAV gi|15234566
AtANT AP2/RAV gi|15235690
AtCBF1 DREB gi|18416562
AtCBF2 DREB gi|18416557
AtCBF3 DREB gi|18416559
AtDREB2A DREB gi|15239107
AtDREB2B DREB gi|15228427
AtERF1 ERF gi|3434967
AtERF2 ERF gi|3434969
AtERF3 ERF gi|3434971
AtERF5 ERF gi|3434975
AtERF6 ERF gi|3298498
AtLEAFY_PETIOLE ERF gi|15240749
AtRAP2.1 DREB gi|18401592
AtRAP2.2 ERF gi|18400321
AtRAP2.3 ERF gi|15228312
AtRAP2.4 DREB gi|2281633
AtRAP2.5 ERF gi|2281635
AtRAP2.6 ERF gi|15218275
AtRAP2.7 AP2/RAV gi|18401775
AtRAP2.10 DREB gi|15234561
AtRAP2.11 AP2/RAV gi|15241182
AtRAP2.12 ERF gi|15220971
AtRAV1 AP2/RAV gi|25091118
AtRAV2 AP2/RAV gi|11357264
AtTINY DREB gi|15239501
Avena sativa AsAP2.1 ERF emb|AM071401 + 1
AsAP2.2 ERF emb|AM071402 + 2
AsAP2.3 ERF emb|AM071403 - 3
AsAP2.4 AP2/RAV emb|AM071404 + 1
AsAP2.5 AP2/RAV emb|AM071405 + 2
AsCBF1 DREB gi|72059436
AsCBF2 DREB gi|72059440
AsCBF3 DREB emb|AM071408 + 2
AsCBF4 DREB gi|72059443
AsERF1 AP2/RAV emb|AM071410 + 2
AsERF2 AP2/RAV emb|AM071411 + 2
Brassica napus BnCBF1 DREB gi|17352283
BnCBF2 DREB gi|17352285
Catharanthus roseus CrORAC1 DREB gi|8346773
CrORAC2 ERF gi|8346775
CrORAC3 ERF gi|8980315
Hordeum vulgare HvCBF1 DREB gi|12658319
HvCBF2 DREB gi|20152903
HvCBF3 DREB gi|12658321
Nicotiana tabacum NtACRE111A DREB gi|12003382
NtACRE111B DREB gi|12003384
NtTSI1 ERF gi|3065895
Lycopersicon esculentum LeCBF1 DREB gi|18535580
Oryza sativa OsDREB1A DREB gi|22594969
OsDREB1B DREB gi|22594973
OsDREB2 DREB gi|22594971
Secale cereale ScCBF DREB gi|17148651
Triticum aestivum TaDREB DREB gi|17226801
1) Nucleotide sequences that have been used in the phylogenetic analysis or in the ClustalW alignment have been translated in to amino acid sequences using the indicated reading frames.
==== Refs
Kellogg EA Relationships of cereal crops and other grasses Proc Natl Acad Sci U S A 1998 95 2005 2010 9482825 10.1073/pnas.95.5.2005
Bennett MD Smith JB Nuclear dna amounts in angiosperms Philos Trans R Soc Lond B Biol Sci 1976 274 227 274 6977
Green C Oats in a new era. 1999 Semundo Limited , 49 North Road, Great Abington, Cambridge 88
Glore SR Van Treeck D Knehans AW Guild M Soluble fiber and serum lipids: a literature review J Am Diet Assoc 1994 94 425 436 8144811 10.1016/0002-8223(94)90099-X
Brown L Rosner B Willett WW Sacks FM Cholesterol-lowering effects of dietary fiber: a meta-analysis Am J Clin Nutr 1999 69 30 42 9925120
Trowell HC Burkitt DP Western diseases: their emergence and prevention. Cambridge, MA: Harvard University Press 1981
Janatuinen EK Pikkarainen PH Kemppainen TA Kosma VM Jarvinen RM Uusitupa MI Julkunen RJ A comparison of diets with and without oats in adults with coeliac disease New Engl J Med 1995 333 1033 – 1037 7675045 10.1056/NEJM199510193331602
Janatuinen EK Kemppainen TA Julkunen RJ Kosma VM Maki M Heikkinen M Uusitupa MI No harm from five year ingestion of oats in coeliac disease. Gut 2002 50 332 335 11839710 10.1136/gut.50.3.332
Størsrud S Olsson M Arvidsson-Lenner1 R. N L.A. Nilsson O Kilander A Adult coeliac patients do tolerate large amounts of oats. European Journal of Clinical Nutrition 2003 57 163 169 12548312 10.1038/sj.ejcn.1601525
Duve KJ White PJ Extraction and identification of antioxidants in oats. Journal of the American Oil Chemists Society 1991 68 365–370
Forssell P Cetin M Wirtanen G Malkki Y Antioxidative effects of oat oil and its fractions. Fett Wissenschaft Technologie-Fat Science Technology 1990 92 319–321
Xing Y White PJ Identification and function of antioxidants from oat groats and hulls J Am Oil Chem Soc 1997 74 303 307
Auerbach RH Gray DA Oat antioxidant extraction and measurement—towards a commercial process. Journal of the Science of Food and Agriculture 1999 79 273–282 10.1002/(SICI)1097-0010(19990301)79:3<385::AID-JSFA260>3.0.CO;2-L
Tian LL White PJ Antioxidant activity of oat extract in soybean and cottonseed oils. J Am Oil Chem Soc 1994 70 1079 11085
Thomashow MF Molecular genetics of cold acclimation in higher plants Advances in genetics 1990 28 99 131
Sarhan F Danyluk J Engineering cold-tolerant crops -throwing the master switch Trends in plant science 1998 3 289 291 10.1016/S1360-1385(98)01285-0
Danyluk J Houde M Rassart E Sarhan F Differential expression of a gene encoding an acidic dehydrin in chilling sensitive and freezing tolerant gramineae species FEBS Lett 1994 344 20 24 7910142 10.1016/0014-5793(94)00353-X
Tatusov RL Fedorova ND Jackson JD Jacobs AR Kiryutin B Koonin EV Krylov DM Mazumder R Mekhedov SL Nikolskaya AN Rao BS Smirnov S Sverdlov AV Vasudevan S Wolf YI Yin JJ Natale DA The COG database: an updated version includes eukaryotes BMC Bioinformatics 2003 4 41 12969510 10.1186/1471-2105-4-41
NCBI Expressed Sequence Tags Database [http://www.ncbi.nlm.nih.gov/dbEST] 2005
Tsuda K Tsvetanov S Takumi S Mori N Atanassov A Nakamura C New members of a cold-responsive group-3 Lea/Rab-related Cor gene family from common wheat (Triticum aestivum L.) Genes Genet Syst 2000 75 179 188 11126566 10.1266/ggs.75.179
Jang JY Kim DG Kim YO Kim JS Kang H An expression analysis of a gene family encoding plasma membrane aquaporins in response to abiotic stresses in Arabidopsis thaliana Plant Mol Biol 2004 54 713 725 15356390 10.1023/B:PLAN.0000040900.61345.a6
Moons A Gielen J Vandekerckhove J Van der Straeten D Gheysen G Van Montagu M An abscisic-acid- and salt-stress-responsive rice cDNA from a novel plant gene family Planta 1997 202 443 454 9265787 10.1007/s004250050148
Chinnusamy V Ohta M Kanrar S Lee BH Hong X Agarwal M Zhu JK ICE1: a regulator of cold-induced transcriptome and freezing tolerance in Arabidopsis Genes Dev 2003 17 1043 1054 12672693 10.1101/gad.1077503
Seki M Narusaka M Ishida J Nanjo T Fujita M Oono Y Kamiya A Nakajima M Enju A Sakurai T Satou M Akiyama K Taji T Yamaguchi-Shinozaki K Carninci P Kawai J Hayashizaki Y Shinozaki K Monitoring the expression profiles of 7000 Arabidopsis genes under drought, cold and high-salinity stresses using a full-length cDNA microarray Plant J 2002 31 279 292 12164808 10.1046/j.1365-313X.2002.01359.x
Fowler S Thomashow MF Arabidopsis transcriptome profiling indicates that multiple regulatory pathways are activated during cold acclimation in addition to the CBF cold response pathway Plant Cell 2002 14 1675 1690 12172015 10.1105/tpc.003483
Thomashow MF Plant Cold Acclimation: Freezing Tolerance Genes and Regulatory Mechanisms Annu Rev Plant Physiol Plant Mol Biol 1999 50 571 599 15012220 10.1146/annurev.arplant.50.1.571
Sakuma Y Liu Q Dubouzet JG Abe H Shinozaki K Yamaguchi-Shinozaki K DNA-binding specificity of the ERF/AP2 domain of Arabidopsis DREBs, transcription factors involved in dehydration- and cold-inducible gene expression Biochem Biophys Res Commun 2002 290 998 1009 11798174 10.1006/bbrc.2001.6299
Dubouzet JG Sakuma Y Ito Y Kasuga M Dubouzet EG Miura S Seki M Shinozaki K Yamaguchi-Shinozaki K OsDREB genes in rice, Oryza sativa L., encode transcription activators that function in drought-, high-salt- and cold-responsive gene expression Plant J 2003 33 751 763 12609047 10.1046/j.1365-313X.2003.01661.x
Cardle L Ramsay L Milbourne D Macaulay M Marshall D Waugh R Computational and experimental characterization of physically clustered simple sequence repeats in plants Genetics 2000 156 847 854 11014830
Stockinger EJ Gilmour SJ Thomashow MF Arabidopsis thaliana CBF1 encodes an AP2 domain-containing transcriptional activator that binds to the C-repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in response to low temperature and water deficit Proc Natl Acad Sci U S A 1997 94 1035 1040 9023378 10.1073/pnas.94.3.1035
Gilmour SJ Zarka DG Stockinger EJ Salazar MP Houghton JM Thomashow MF Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional activators as an early step in cold-induced COR gene expression Plant J 1998 16 433 442 9881163 10.1046/j.1365-313x.1998.00310.x
Weber JL Informativeness of human (dC-dA)n.(dG-dT)n polymorphisms. Genomics 1990 7 524 5530 1974878 10.1016/0888-7543(90)90195-Z
Hamann A Zink D Nagl W Microsatellite fingerprinting in the genus Phaseolus. Genome 1995 38 507 5515
Morgante M Olivieri AM PCR-amplified microsatellites as markers in plant genetics Plant J 1993 3 175 182 8401603 10.1046/j.1365-313X.1993.t01-9-00999.x
Rongwen J Akkaya MS Bhahwat AA Lavi U Cregan PB The use of microsatellite DNA markers for soybean genotype identification. Theoretical and Applied Genetics 1995 90 43 448 10.1007/BF00220994
Sharon D Adato A Mhameed S Lavi U Hillel J Gomolka M Epplen C Epplen J DNA fingerprints in plants using simple-sequence repeat and minisatellite probes. HortScience 1995 30 109 1112
O'Donoughue LS Wang Z Roder M Kneen B M. L Sorrells ME Tanksley SD An RFLP-based map of oat on a cross between two diploid taxa (Avena atlantica x A. hirtula). Genome 1992 35 765 771
Rayapati PJ Gregory JW Lee M Wise RP A linkage map of diploid oat Avena based on RFLP loci and a locus conferring resistance to Puccinia coronata var. avenae. Theoretical and Applied Genetics 1995 89 831 837
O'Donoughue LS Kianian SF Rayapati PJ Penner GA Sorrells ME Tanksley SD Phillips RL Rines HW Lee M Fedak G Molnar SJ Hoffman D Salas CA Wu B Autrique E Van Deynze A A molecular map of cultivated oat. Genome 1995 38 368 380
Wight CP Tinker NA Kianian SF Sorrells ME O'Donoughue LS Hoffman DL Groh S Scoles GJ Li CD Webster FH Phillips RL Rines HW Livingston SM Armstrong KC Fedak G Molnar SJ A molecular marker map in 'Kanota' x 'Ogle' hexaploid oat (Avena spp.) enhanced by additional markers and a robust framework Genome 2003 46 28 47 12669794 10.1139/g02-099
Gharti-Chhetri G Olsson O Establishment of a highly efficient callus proliferation and plant regeneration systems from different explants of seven commercial Swedish oat (Avena sativa L.) cultivars. (Submitted to Plant Cell Tissue and Organ Culture) 2005
Olsson O Gharti-Chhetri GB Novel transformation method and transformed plants Patent pending. 2005
Ausubel FM Brent R Kingston RE Moore DD Seidman JG Smith JA Struhl K Short Protocols in molecular Biology 1995 New York , John Wiley & Sons
Church GM Gilbert W The genomic sequencing technique Prog Clin Biol Res 1985 177 17 21 2989835
Yulan P Naomi T Meng KL Minoru SHK Total RNAs by a Universal PCR Amplification Method. Genome Research 2001 11 1553 1558 11544199 10.1101/gr.185501
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
NCBI Basic Local Alignment Search Tool [http://www.ncbi.nlm.nih.gov/blast] 2005
Rice P Longden I Bleasby A EMBOSS: The European Molecular Biology Open Software Suite Trends in Genetics 2000 16 276 277 10827456
EMBOSS The European Molecular Biology Open Software Suite [http://emboss.sourceforge.net] 2005
Zdobnov EM Apweiler R InterProScan - an integration platform for the signature-recognition methods in InterPro Bioinformatics 2001 17 847 848 11590104 10.1093/bioinformatics/17.9.847
EBI InterProScan [http://www.ebi.ac.uk/InterProScan/] 2005
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 Research 1994 22 4673 4680 7984417
NCBI National Center for Biotechnology Information [http://www.ncbi.nlm.nih.gov] 2005
Huang X Madan A CAP3: A DNA sequence assembly program Genome Res 1999 9 868 877 10508846 10.1101/gr.9.9.868
TIGR TGI Clustering tools [http://www.tigr.org/software/] 2003
Zhang Z Schwartz S Wagner L Miller W A greedy algorithm for aligning DNA sequences J Comput Biol 2000 7 203 214 10890397 10.1089/10665270050081478
Abajian C Sputnik [http://espressosoftware.com/pages/sputnik.jsp] 1994
EBI EMBL-Nucleotide Sequence Database [http://www.ebi.ac.uk/embl/index.html] 2005
|
16137325
|
PMC1236939
|
CC BY
|
2021-01-04 16:03:52
|
no
|
BMC Plant Biol. 2005 Sep 1; 5:18
|
utf-8
|
BMC Plant Biol
| 2,005 |
10.1186/1471-2229-5-18
|
oa_comm
|
==== Front
BMC Vet ResBMC Veterinary Research1746-6148BioMed Central London 1746-6148-1-21622129710.1186/1746-6148-1-2Research ArticleOral infection with the Salmonella enterica serovar Gallinarum 9R attenuated live vaccine as a model to characterise immunity to fowl typhoid in the chicken Wigley Paul [email protected] Scott [email protected] Claire [email protected] Richard [email protected] Adrian [email protected] Paul [email protected] Department of Veterinary Pathology, University of Liverpool, Leahurst, Neston, CH64 7TE, Merseyside, UK 2 Institute for Animal Health, Compton Laboratory, Compton, Newbury, UK, RG20 7NN, Berkshire, UK2005 12 9 2005 1 2 2 19 7 2005 12 9 2005 Copyright © 2005 Wigley et al; licensee BioMed Central Ltd.2005Wigley 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
Salmonella enterica serovar Gallinarum (S. Gallinarum) is the causative agent of fowl typhoid, a severe systemic disease of chickens that results in high mortality amongst infected flocks. Due to its virulence, the immune response to S. Gallinarum is poorly characterised. In this study we have utilised infection by the live attenuated S. Gallinarum 9R vaccine strain in inbred chickens to characterise humoral, cellular and cytokine responses to systemic salmonellosis.
Results
Infection with 9R results in a mild systemic infection. Bacterial clearance at three weeks post infection coincides with increases in circulating anti-Salmonella antibodies, increased T cell proliferation to Salmonella challenge and increased expression of interferon gamma. These responses peak at four weeks post infection, then decline. Only modest increases of expression of the pro-inflammatory cytokine interleukin-1β were detected early in the infection.
Conclusion
Infection of chickens with the 9R vaccine strain induces a mild form of systemic salmonellosis. This induces both cellular and humoral immune responses, which peak soon after bacterial clearance. Unlike enteric-associated Salmonella infections the immune response is not prolonged, reflecting the absence of persistence of Salmonella in the gastrointestinal tract. The findings here indicate that the use of the S. Gallinarum 9R vaccine strain is an effective model to study immunity to systemic salmonellosis in the chicken and may be employed in further studies to determine which components of the immune response are needed for protection.
==== Body
Background
Salmonella enterica serovar Gallinarum (S. Gallinarum) is the causative agent of fowl typhoid, a severe systemic disease of chickens and other galliforme birds [1]. S. Gallinarum is a non-motile Gram negative rod and along with the closely related Salmonella enterica serovar Pullorum is host-specific for poultry, but rarely, if ever, presents a risk of zoonotic transmission to man. Infection in chickens may occur at all ages and is typified by severe hepatosplenomegaly accompanied by characteristic liver 'bronzing', anaemia and septicaemia [1]. S. Gallinarum is primarily associated with the mononuclear phagocyte system and resides primarily within macrophages in the liver and spleen [2,3]. It is only found in the gastrointestinal tract early in the infection, usually through faecal-oral transmission, and in the end stage of fowl typhoid where bacteria are shed into the intestines leading to substantial haemorrhaging of the intestinal wall [3]. Infection leads to high rates of morbidity and mortality with a recent study describing a mortality rate in excess of 60% in experimentally infected outbred chickens [4]. Although control programmes including vaccination have largely controlled the disease in Europe and North America, it remains of high economic importance to developing poultry industries in Asia and South America.
The high mortality and morbidity rates associated with S. Gallinarum make effective study of immunity to infection difficult to achieve. Although studies in inbred genetically resistant and susceptible chickens have demonstrated the role of the innate immune system to some extent [3], the role of the adaptive immune response and in particular cellular responses has not been described other than a few serological studies [1,2,5]. In order to further characterise the immune response to S. Gallinarum, we utilised an infection model with the live attenuated fowl typhoid vaccine 9R. The 9R vaccine strain developed in the 1950s, has a 'semi-rough' lipopolysaccharide structure, but the nature of its attenuation is not known [6,7]. Although highly attenuated compared to its parental strain S. Gallinarum 9, the 9R vaccine strain still results in systemic disease with pathology in the liver and spleen, and bacterial persistence for several weeks at these sites [5]. Therefore although the vaccine strain does not cause significant mortality, it causes a mild form of systemic salmonellosis. This allows more detailed study of the immune response to be undertaken without using high numbers of animals needed to determine immune responses associated with clearance due to high mortality rates.
In this study we have determined both humoral and cellular immune responses to systemic salmonellosis in an inbred chicken line, and investigate expression of two key cytokines. Line 72 is derived from White Leghorn Chickens and is moderately susceptible to systemic salmonellosis [3]. Previous studies have determined both cellular and humoral responses to Salmonella enterica serovar Typhimurium in the chicken [8,9]. Although S. Typhimurium is mainly associated with the gastrointestinal tract in chickens, it causes a transient primary systemic infection. Control of this systemic infection appears to be dependent on cell-mediated immunity as clearance of bacteria from the spleen and liver coincides with the height of T-cell proliferative activity and expression of the T helper 1(Th1)-type cytokine interferon-γ (IFN-γ) [8,9]. Infection with S. Typhimurium also results in specific IgG, IgM and IgA antibody responses [8], as does vaccination with killed Salmonella vaccines [10]. Infection with either Salmonella serovar Typhimurium or serovar Enteritidis leads to prolonged high titres of specific antibody, probably as a consequence of the prolonged persistence of these serovars in the gastrointestinal tract, a phenomenon that is not found with S. Gallinarum [9,11-13] Whilst it is clear specific antibody protects against secondary systemic infection [10], its role in clearance of primary infection is less clear. Following bursectomy, depletion of the Bursa of Fabricius the primary lymphoid organ associated with B lymphocyte development in the chicken, [14,15], conflicting results have been found on the role of antibody in Salmonella clearance. Therefore the role of antibody is not yet clearly defined in primary avian Salmonella infection.
In this study we have initiated investigation of immunity to systemic infection with S. Gallinarum. We have determined both the specific anti-S. Gallinarum antibody (IgG and IgM) response, the cellular response and expression of the key cytokines IFN-γ and Interleukin-1β (IL-1β) following oral infection of inbred chickens with S. Gallinarum 9R.
Results
The S. Gallinarum 9R vaccine strain causes a low level systemic infection
Following infection S. Gallinarum 9R was detected in the spleen and liver at one and two weeks post infection (Table 1), though no bacteria were detected by three weeks post infection or at any subsequent time point. At both one and two weeks post infection inflammatory signs including mild to moderate hepatosplenomegaly were detected in infected birds. This level of pathology is consistent with previous studies of the 9R strain [5].
Table 1 Counts of S. Gallinarum 9R vaccine strain from tissues following oral inoculation of three-week old Line 72 chickens (±SEM)
Tissue Weeks post infection
1 2 3
Mean count Log10 cfu/g Number positive Mean count Log10 cfu/g Number positive Mean count Log10 cfu/g Number positive
Liver 1.84 ± 0.37 5/5 1.43 ± 0.29 5/5 <1 0/5
Spleen 1.39 ± 0.57 4/5 < 1 3/5* <1 0/5
*detected only after enrichment in selenite broth
Cellular and humoral responses to experimental infection with 9R
IgM antibody responses were detected at one week post infection, and subsequently decreased (Figure 1a), whereas IgG responses reached a peak three weeks post infection and subsequently declined (Figure 1b). Cellular responses followed a similar pattern to IgG production with significantly higher levels of T-cell proliferation (P < 0.05) to Salmonella antigen in the infected over control groups found at three-to-four weeks post infection (Figure 2). These findings indicate that oral infection with 9R elicits both humoral and cellular immune responses that coincide with clearance of the bacterium from the liver and spleen.
Figure 1 Antibody responses to S. Gallinarum 9R. Serum antibody responses to S. Gallinarum lysate antigen in S. Gallinarum 9R infected and control Line 72 chickens infected orally at three weeks of age as determined by ELISA (±SEM). (A) IgM response (serum diluted 1:50) and (B) IgG response (serum diluted 1:200)
Figure 2 Antigen-specific T lymphocyte proliferation to Salmonella. Antigen-specific proliferation of splenic lymphocytes from Line 72 chickens infected with S. Gallinarum 9R or infected controls (±SEM). Proliferation was determined by the uptake of tritiated thymidine measured 48 h after culture of cells with soluble S. Gallinarum antigen. Differences between control and infected groups were analysed by ANOVA indicating significantly increased proliferation in infected over control birds at 3 and 4 weeks post infection (P < 0.05)
Cytokine expression following S. Gallinarum 9R infection
Expression of the Th1-type cytokine IFN-γ followed a similar pattern to that of T-cell proliferation with infected birds showing increased expression over the controls at three weeks (P = 0.03) and four weeks (P < 0.01) post infection with up to eight-fold increases in expression found (Figure 3). The increase is consistent with T-lymphocyte activation, particularly Th1 responses. Only modest increases of expression of the pro-inflammatory cytokine IL-1β were found following infection, though significantly greater in infected birds at two weeks post infection (P = 0.02). This increase corresponds to presence of bacteria in the spleen and mild hepatosplenomegaly in the infected group. Previous studies have suggested that presence of high bacterial numbers in the spleen induces high expression levels of IL-1β [9]. However in this study relatively limited inflammation was found and it seems likely that 9R induces only a limited pro-inflammatory response.
Figure 3 Expression of IFN-γ and IL1-β following infection with S. Gallinarum 9R. Expression of the cytokines IFN-γ (A) and IL1-β (B) in the spleen of Line 72 chickens infected with S. Gallinarum 9R (±SEM). Expression was determined by quantitative reverse-transcription real-time PCR. Data are displayed as the mean fold-change in expression in infected birds (n = 5) in comparison to uninfected controls (n = 5) at the same time point. Differences between control and infected groups were analysed by ANOVA indicating significantly increased expression of IFN-γ in infected over control birds at 3 and 4 weeks post infection (P < 0.05).
Discussion
The data presented here indicate that systemic Salmonella infection of the chicken induces both cellular and humoral responses in the chicken. Both responses peak at three to four weeks post infection, a point that coincides with bacterial clearance. Previous studies of immune responses in avian salmonellosis have concentrated on virulent or attenuated strains of S. Typhimurium and S. Enteritidis [9,16-19]. The responses found in these studies also demonstrated both humoral and cellular responses that peak at similar times to those described here, though in contrast to this study, these infections elicit responses that remain considerably higher than in this study, where responses declined rapidly after four weeks post infection. Both S. Typhimurium and S. Enteritidis may persist in the gastrointestinal tract for many weeks, whereas S. Gallinarum infection generally results either in the mortality of susceptible birds or bacterial clearance in resistant birds within three to four weeks of initial infection, although occasionally persistent infection occurs [3,20]. It seems that bacterial persistence in the gastrointestinal tract maintains a more prolonged immune response. In addition persistent, low-level, systemic infection of chickens by Salmonella enterica serovar Pullorum leads to prolonged high titre antibody responses and T lymphocyte proliferation [21].
Although the duration of the primary immune response is shorter in S. Gallinarum infection, as with the early stages of S. Typhimurium infection in the chicken, clearance from the spleen and liver coincides with increased T lymphocyte proliferation and expression of IFN-γ [9]. Comparative infection studies of Salmonella-resistant naked neck chicken breed and commercial layers have also shown correlation between the cellular immune response and protection against fowl typhoid [22]. In many ways the biology of S. Gallinarum infection is more akin to S. Typhimurium infection of the mouse than the chicken. In common with S. Gallinarum in the chicken, murine infection with S. Typhimurium results in a severe systemic 'typhoid-like' infection. The development of the murine immune response to S. Typhimurium has been well characterised and has recently been reviewed [23]. In the early stages of both murine and avian systemic salmonellosis, bacterial numbers are initially controlled through the innate immune system, and in particular through the generation of reactive oxygen intermediates [3,23]. In the mouse initiation of the adaptive immunity relies on the action of a number of cytokines including IFN-γ, Interleukin-12 (IL-12), Interluekin-18 (IL-18) and tumour necrosis factor-α. Production of IL-12 and IL-18, primarily by macrophages leads to expression of IFN-γ by natural killer cells and T lymphocytes, which in turn leads to increased macrophage activation. These, primarily Th1, cytokine responses lead to the development of CD4+ T-cell responses that lead to clearance of Salmonella from the tissues. The response to S. Galllinarum 9R mirrors this, with an increase in IFN-γ expression correlating to increased T-cell proliferation and clearance of the vaccine strain from the tissues. Antibody responses, initially IgM followed by IgG are produced to 9R in a classical primary response. These also correlate to the clearance of Salmonella. The relative roles of humoral and cellular immunity to clearance are not yet known, though as S. Gallinarum are believed to survive and multiply within macrophages [2-4], it would appear to be more likely that Th1-mediated cellular responses are more important in clearance. Further functional studies will be used to determine the relative roles of cellular and humoral responses in protection.
The infection model used in this study is a useful tool in studying the immune response to S. Gallinarum, but without the high mortality or morbidity rates found with virulent strains, even in genetically resistant birds. Similar attenuated vaccines strains have been used extensively in the S. Typhimurium murine model [23] and also to characterise the immune responses to Salmonella infection in cattle [24]. The data presented here indicates that an immune response consistent with that seen in other animal models of systemic salmonellosis is elicited following oral infection with the 9R vaccine strain. This represents a valuable model to study immunity to fowl typhoid in the chicken. The studies presented here indicate that infection generates both humoral and a Th1-mediated cellular response. Further studies with this model will allow the determination of which components of the response are protective, aiding future rational avian Salmonella vaccine design.
Conclusion
The S. Gallinarum 9R vaccine strain is a suitable model to study the immune response to systemic salmonellosis in the chicken. Infection with 9R induces both antibody and Th1-type T cell responses that are associated with bacterial clearance.
Methods
Experimental animals
Specific pathogen-free (SPF) Line 72 inbred White Leghorn chickens were obtained form the Poultry Production Unit, Institute for Animal Health, Compton, UK. Birds were reared on wire cages initially at an ambient temperature of 30°C then at 21°C from three weeks old. Birds were given ad libitum access to water and a vegetable protein based diet (SDS, Witham, Essex, UK). All experimental work involving animals was performed under the conditions of a Home Office project licence and of the local ethics committee meeting the requirements of UK legislation.
Bacterial strains
The S. Gallinarum 9R vaccine strain [6] was cultured in Luria Bertani (LB) broth (Difco, Becton-Dickinson Labware, Cowley, Oxford, UK) at 37°C in an orbital shaking incubator at 150 rpm from stocks held at -70°C in LB broth supplemented with 30% glycerol.
Production of Salmonella lysate antigen
A soluble protein antigen lysate preparation was prepared from S. Gallinarum 9, the virulent strain from which the 9R vaccine was produced, as previously described [9]. The antigen preparation was subsequently used for ELISA and T-cell proliferation assays
Experimental infection with the 9R vaccine strain
Fifty 3 week-old Line 72 chickens were divided into two groups of equal size and housed separately as described above. Prior to infection five birds from each group were bled from the wing vein to obtain serum. The birds of one group were then infected orally with 108 CFU of the S. Gallinarum 9R vaccine strain in a volume of 0.3 ml of LB broth. The second group remained uninfected as controls. At 1, 2, 3, 4 and 6 weeks post infection, five birds from each group were killed for post mortem analysis. At each time point samples of spleen and liver were taken aseptically for bacteriological analysis. A section of splenic tissue was taken into RPMI1640 containing 100 U/ml penicillin, 1 μg/ml streptomycin and 5% bovine serum to isolate splenocytes for T-cell proliferation assays. A small sample of splenic tissue was also obtained for isolation of RNA using RNAlater (Qiagen, Crawley, UK) to protect against any degradation. Birds were also bled by cardiac puncture to obtain serum. For bacteriological analysis samples were homogenised in sterile phosphate buffered saline (PBS) using Griffith's tube homogenisers, then serially diluted in PBS and plated onto Brilliant Green Agar (Difco, Becton-Dickinson Labware). Plates were then incubated at 37°C for 24 h, then the bacterial count determined. Samples were also enriched by adding an equal volume of double strength selenite broth, followed by overnight incubation at 37°C. Enriched samples were plated onto Brilliant Green Agar and incubated as described above. Growth was then recorded as Salmonella positive or negative.
Measurement of anti-Salmonella antibody responses by ELISA
Anti-Salmonella IgG and IgM responses were determined by ELISA on pre-infection and post mortem serum samples using plates coated with S. Gallinarum lysate antigen as described previously for S. Pullorum and S. Typhimurium [9,25].
T-cell proliferation assay
Single cell suspensions of splenocytes were prepared from post mortem samples by passing splenic tissue through Falcon cell strainers (Life Technologies, Paisley, UK) in RPMI1640 containing 100 U/ml penicillin, 1 μg/ml streptomycin and 5% bovine serum. The majority of erythrocytes were removed by centrifugation at 35 × g for 10 minutes. Cell proliferation to S. Gallinarum antigen was determined by uptake of tritiated thymidine as described previously [9].
Quantitative analysis of cytokine mRNA expression
Total RNA was isolated from RNAlater-protected samples using RNeasy min kits (Qiagen, Crawley, UK) following manufacturer's protocols. Levels of expression of the cytokines IL-1β and IFN-γ were determined by real-time reverse transcription-polymerase chain reaction (RT-PCR) using the ABI Prism 7700 Sequence Detection System (TaqMan®; PE Applied Biosystems, Warrington, UK) as previously described [9,26-28]. Values for expression of mRNA were corrected against the expression of 28S rRNA as a 'housekeeping' gene.
Statistical analysis
Statistical analysis was performed either using Microsoft Excel or Minitab for Windows. Comparison between infected and control groups was made by ANOVA. Values of P < 0.05 were taken as significant.
Authors' contributions
PW conceived the experimental outline, conducted the in vivo experiments, cell proliferation assays, analysed the data and co-wrote the manuscript, SH performed the cytokine expression and assisted with the bacteriology. CP assisted with in vivo experiments and performed the ELISA assays, RB developed and assisted with the proliferation assays, AS helped in experimental design and preparation of the manuscript, PB co-designed the experiments and co-wrote the manuscript.
Acknowledgements
The authors wish to thank the Biotechnology and Biological Sciences Research Council for funding.
==== Refs
Shivaprasad HL Fowl typhoid and pullorum disease Rev Sci Tech 2000 19 405 424 10935271
Barrow PA Huggins MB Lovell MA Host specificity of Salmonella infection in chickens and mice is expressed in vivo primarily at the level of the reticuloendothelial system Infect Immun 1994 62 4602 4610 7927727
Wigley P Hulme SD Bumstead N Barrow PA In vivo and in vitro studies of genetic resistance to systemic salmonellosis in the chicken encoded by the SAL1 locus Microbes Infect 2002 4 1111 1120 12361910 10.1016/S1286-4579(02)01635-0
Jones MA Wigley P Page KL Hulme SD Barrow PA Salmonella enterica serovar Gallinarum requires the Salmonella pathogenicity island 2 type III secretion system but not the Salmonella pathogenicity island 1 type III secretion system for virulence in chickens Infect Immun 2001 69 5471 5476 11500419 10.1128/IAI.69.9.5471-5476.2001
Silva EN Snoeyenbos GH Weinack OM Smyser CF Studies on the use of 9R strain of Salmonella gallinarum as a vaccine in chickens Avian Dis 1981 25 38 52 7271663
Smith HW The use of live vaccines in experimental Salmonella gallinarum infection in chickens with observations on their interference effect J Hyg (Lond) 1956 54 419 432 13367409
Smith HW The immunity to Salmonella gallinarum infection in chickens produced by live cultures of members of the Salmonella genus J Hyg (Lond) 1956 54 433 439 13367410
Beal RKWPPCHSDBPASAL Age at primary infection with Salmonella enterica serovar Typhimurium in the chicken influences persistence of infection and subsequent immunity to re-challenge Vet Immunol Immunopathol 2004 100 151 164 15207453 10.1016/j.vetimm.2004.04.005
Beal RK Powers C Wigley P Barrow PA Smith AL Temporal dynamics of the cellular, humoral and cytokine responses in chickens during primary and secondary infection with Salmonella enterica serovar Typhimurium Avian Pathol 2004 33 25 33 14681065 10.1080/03079450310001636282
Woodward MJ Gettinby G Breslin MF Corkish JD Houghton S The efficacy of Salenvac, a Salmonella enterica subsp. Enterica serotype Enteritidis iron-restricted bacterin vaccine, in laying chickens Avian Pathol 2002 31 383 392 12396340 10.1080/03079450220141660
Barrow P Serological analysis for antibodies to S enteritidis Vet Rec 1991 128 43 44 2017849
Brito JR Hinton M Stokes CR Pearson GR The humoral and cell mediated immune response of young chicks to Salmonella typhimurium and S. Kedougou Br Vet J 1993 149 225 234 8334505
Cooper GL Venables LM Woodward MJ Hormaeche CE Vaccination of chickens with strain CVL30, a genetically defined Salmonella enteritidis aroA live oral vaccine candidate Infect Immun 1994 62 4747 4754 7927750
Corrier DE Elissalde MH Ziprin RL DeLoach JR Effect of immunosuppression with cyclophosphamide, cyclosporin, or dexamethasone on Salmonella colonization of broiler chicks Avian Dis 1991 35 40 45 2029260
Desmidt M Ducatelle R Mast J Goddeeris BM Kaspers B Haesebrouck F Role of the humoral immune system in Salmonella enteritidis phage type four infection in chickens Vet Immunol Immunopathol 1998 63 355 367 9656424 10.1016/S0165-2427(98)00112-3
Babu U Dalloul RA Okamura M Lillehoj HS Xie H Raybourne RB Gaines D Heckert RA Salmonella enteritidis clearance and immune responses in chickens following Salmonella vaccination and challenge Vet Immunol Immunopathol 2004 101 251 257 15350755 10.1016/j.vetimm.2004.05.002
Babu U Scott M Myers MJ Okamura M Gaines D Yancy HF Lillehoj H Heckert RA Raybourne RB Effects of live attenuated and killed Salmonella vaccine on T-lymphocyte mediated immunity in laying hens Vet Immunol Immunopathol 2003 91 39 44 12507848 10.1016/S0165-2427(02)00265-9
Berndt A Methner U Gamma/delta T cell response of chickens after oral administration of attenuated and non-attenuated Salmonella typhimurium strains Vet Immunol Immunopathol 2001 78 143 161 11182154 10.1016/S0165-2427(00)00264-6
Dueger EL House JK Heithoff DM Mahan MJ Salmonella DNA adenine methylase mutants elicit protective immune responses to homologous and heterologous serovars in chickens Infect Immun 2001 69 7950 7954 11705984 10.1128/IAI.69.12.7950-7954.2001
Wigley P Genetic resistance to Salmonella infection in domestic animals Res Vet Sci 2004 76 165 169 15046948 10.1016/S0034-5288(03)00117-6
Wigley P Hulme SD Powers C Beal RK Berchieri AJ Smith A Barrow P Infection of the Reproductive Tract and Eggs with Salmonella enterica Serovar Pullorum in the Chicken Is Associated with Suppression of Cellular Immunity at Sexual Maturity Infect Immun 2005 73 2986 2990 15845505 10.1128/IAI.73.5.2986-2990.2005
Alvarez MT Ledesma N Tellez G Molinari JL Tato P Comparison of the immune responses against Salmonella enterica serovar Gallinarum infection between naked neck chickens and a commercial chicken line Avian Pathol 2003 32 193 203 12745372 10.1080/0307945021000071605
Mastroeni P Menager N Development of acquired immunity to Salmonella J Med Microbiol 2003 52 453 459 12748263 10.1099/jmm.0.05173-0
Villarreal-Ramos B Manser J Collins RA Dougan G Chatfield SN Howard C Immune responses in calves immunised orally or subcutaneously with a live Salmonella typhimurium aro vaccine. Vaccine 1998 16 45 54 9607008 10.1016/S0264-410X(97)00156-4
Wigley P Berchieri AJ Page KL Smith AL Barrow PA Salmonella enterica serovar Pullorum persists in splenic macrophages and in the reproductive tract during persistent, disease-free carriage in chickens Infect Immun 2001 69 7873 7879 11705970 10.1128/IAI.69.12.7873-7879.2001
Kaiser P Rothwell L Galyov EE Barrow PA Burnside J Wigley P Differential cytokine expression in avian cells in response to invasion by Salmonella typhimurium, Salmonella enteritidis and Salmonella gallinarum Microbiology 2000 146 Pt 12 3217 3226 11101679
Kaiser P Underwood G Davison F Differential cytokine responses following Marek's disease virus infection of chickens differing in resistance to Marek's disease J Virol 2003 77 762 768 12477883 10.1128/JVI.77.1.762-768.2003
Withanage GS Kaiser P Wigley P Powers C Mastroeni P Brooks H Barrow P Smith A Maskell D McConnell I Rapid expression of chemokines and proinflammatory cytokines in newly hatched chickens infected with Salmonella enterica serovar typhimurium Infect Immun 2004 72 2152 2159 15039338 10.1128/IAI.72.4.2152-2159.2004
|
16221297
|
PMC1236940
|
CC BY
|
2021-01-04 16:29:46
|
no
|
BMC Vet Res. 2005 Sep 12; 1:2
|
utf-8
|
BMC Vet Res
| 2,005 |
10.1186/1746-6148-1-2
|
oa_comm
|
==== Front
Cardiovasc DiabetolCardiovascular Diabetology1475-2840BioMed Central London 1475-2840-4-141616805210.1186/1475-2840-4-14ReviewDual and pan-peroxisome proliferator-activated receptors (PPAR) co-agonism: the bezafibrate lessons Tenenbaum Alexander [email protected] Michael [email protected] Enrique Z [email protected] Cardiac Rehabilitation Institute, Sheba Medical Center, 52621 Tel-Hashomer, Israel2 Sackler Faculty of Medicine, Tel-Aviv University, 69978 Tel-Aviv, Israel3 Cardiovascular Diabetology Research Foundation, 58484 Holon, Israel2005 16 9 2005 4 14 14 14 9 2005 16 9 2005 Copyright © 2005 Tenenbaum et al; licensee BioMed Central Ltd.2005Tenenbaum 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.
There are three peroxisome proliferator-activated receptors (PPARs) subtypes which are commonly designated PPAR alpha, PPAR gamma and PPAR beta/delta. PPAR alpha activation increases high density lipoprotein (HDL) cholesterol synthesis, stimulates "reverse" cholesterol transport and reduces triglycerides. PPAR gamma activation results in insulin sensitization and antidiabetic action. Until recently, the biological role of PPAR beta/delta remained unclear. However, treatment of obese animals by specific PPAR delta agonists results in normalization of metabolic parameters and reduction of adiposity. Combined treatments with PPAR gamma and alpha agonists may potentially improve insulin resistance and alleviate atherogenic dyslipidemia, whereas PPAR delta properties may prevent the development of overweight which typically accompanies "pure" PPAR gamma ligands. The new generation of dual-action PPARs – the glitazars, which target PPAR-gamma and PPAR-alpha (like muraglitazar and tesaglitazar) are on deck in late-stage clinical trials and may be effective in reducing cardiovascular risk, but their long-term clinical effects are still unknown. A number of glitazars have presented problems at a late stage of clinical trials because of serious side-effects (including ragaglitazar and farglitazar). The old and well known lipid-lowering fibric acid derivative bezafibrate is the first clinically tested pan – (alpha, beta/delta, gamma) PPAR activator. It is the only pan-PPAR activator with more than a quarter of a century of therapeutic experience with a good safety profile. Therefore, bezafibrate could be considered (indeed, as a "post hoc" understanding) as an "archetype" of a clinically tested pan-PPAR ligand. Bezafibrate leads to considerable raising of HDL cholesterol and reduces triglycerides, improves insulin sensitivity and reduces blood glucose level, significantly lowering the incidence of cardiovascular events and new diabetes in patients with features of metabolic syndrome. Clinical evidences obtained from bezafibrate-based studies strongly support the concept of pan-PPAR therapeutic approach to conditions which comprise the metabolic syndrome. However, from a biochemical point of view, bezafibrate is a PPAR ligand with a relatively low potency. More powerful new compounds with pan-PPAR activity and proven long-term safety should be highly effective in a clinical setting of patients with coexisting relevant lipid and glucose metabolism disorders.
==== Body
Peroxisome proliferator-activated receptors
Peroxisome proliferator-activated receptors (PPARs) are nuclear hormone receptors, i.e. ligand-dependent intracellular proteins that stimulate transcription of specific genes by binding to specific DNA sequences following activation by the appropriate ligand. When activated, the transcription factors exert several functions in development and metabolism. There are three PPAR subtypes which are the products of distinct genes and are commonly designated PPAR alpha, PPAR gamma and PPAR beta/delta, or merely delta [1-4]. The PPARs usually heterodimerize with another nuclear receptor, the 9-cis-retinoic acid receptor (RXR), forming a complex that interacts with specific DNA response elements within promoter regions of target genes. When activated by agonist ligand binding, this heterodimer complex recruits transcription coactivators and regulates the transcription of genes involved in the control of lipid and carbohydrate metabolism [1-4].
PPAR alpha, activated by polyunsaturated fatty acids and fibrates, is implicated in regulation of lipid metabolism, lipoprotein synthesis and metabolism and inflammatory response in liver and other tissues. PPAR alpha is highly expressed in tissues with high fatty acid oxidation (like liver, kidney and heart muscle), in which it controls a comprehensive set of genes that regulate most aspects of lipid catabolism. Like several other nuclear hormone receptors, it heterodimerizes with RXR alpha to form a transcriptionally competent complex [1-3,5]. In addition, PPAR-alpha is expressed in vascular endothelial cells, smooth muscle cells, monocyte/macrophages and T lymphocytes. PPAR alpha activation increase HDL cholesterol synthesis, stimulate "reverse" cholesterol transport and reduce triglycerides [1-3,6].
PPAR gamma plays important roles in the regulation of proliferation and differentiation of several cell types, including adipose cells. It has the ability to bind a variety of small lipophilic compounds derived from both metabolism and nutrition. These ligands, in turn, determine cofactor recruitment to PPAR gamma, regulating the transcription of genes in a variety of complex metabolic pathways. PPAR gamma is highly expressed in adipocytes, where it mediates differentiation, promotes lipid storage, and, as a consequence, is thought to indirectly improve insulin sensitivity and enhance glucose disposal in adipose tissue and skeletal muscle [7-9]. Its activation by drugs of the glitazones (thiazolidinediones) group results in insulin sensitization and antidiabetic action.
Until recently, the biological role of PPAR delta remained unclear. Animal studies revealed that PPAR delta play an important role in the metabolic adaptation of several tissues to environmental changes. Treatment of obese animals by specific PPAR delta agonists results in normalization of metabolic parameters and reduction of adiposity. PPAR delta appeared to be implicated in the regulation of fatty acid burning capacities of skeletal muscle and adipose tissue by controlling the expression of genes involved in fatty acid uptake, beta-oxidation and energy uncoupling. PPAR delta is also implicated in the adaptive metabolic response of skeletal muscle to endurance exercise by controlling the number of oxidative myofibers, inducing so and enhancing fatty acid catabolism in muscular tissue [3,6,10]. Moreover, recent studies revealed that ligand activation of these receptors is associated with improved insulin sensitivity and elevated HDL levels thus demonstrating promising potential for targeting PPAR delta in the treatment of obesity, dyslipidemias and type 2 diabetes [11].
Clinical studies of PPAR ligands
Fibric acid derivatives (fibrates) are PPAR alpha ligands. Fibrates have been used in clinical practice for more than four decades as a class of agents known to decrease triglyceride levels while substantially increasing HDL-cholesterol levels, with a limited but significant additional lowering effect on low density lipoprotein (LDL)-cholesterol levels [5]. In addition to their favorable effects on lipid profiles, evidence is mounting that benefits may also stem from the anti-inflammatory and antiatherosclerotic properties of these drugs [12,13]. Although fibrate trials have reported cardiovascular risk reduction in patients with dyslipidemia, it is evident that the favorable alterations in plasma lipids can only partially explain the reduction in cardiovascular events in these studies. This is particularly evident for high-risk individuals, such as diabetics or patients with insulin resistance who may have more pronounced cardiovascular benefits [5,12-15].
Glitazones are synthetic PPAR gamma ligands with well recognized effects on glucose and lipid metabolism. The clinical use of these PPARgamma agonists in type 2 diabetic patients leads to an improved glycemic control and an inhanced insulin sensitivity, and – at least in animal models – to a protective effect on pancreatic beta-cell function. Glitazones may also have cardiovascular benefits. Animal models of atherosclerosis have shown that these drugs reduce the extent of atherosclerotic lesions and inhibit macrophage accumulation. Clinical studies have also shown that these drugs improve the lipid profile of patients at risk of developing atherosclerosis and reduce circulating levels of inflammatory markers [16-18]. However, they can produce adverse effects, generally mild or moderate, but some of them (mainly peripheral edema and weight gain) may lead to treatment cessation.
Currently, clinical studies regarding PPAR delta ligands are lacking. Given the results obtained with animal models, PPAR delta agonists may have therapeutic usefulness in metabolic syndrome by increasing fatty acid consumption in skeletal muscle and adipose tissue [19]. Probably, weight reduction could be expected as well.
Dual and pan-PPAR co-agonism
Combined treatments with PPAR gamma and alpha agonists may potentially improve insulin resistance and alleviate atherogenic dyslipidemia, whereas PPAR delta properties may prevent the development of overweight which typically accompanies "pure" PPAR gamma ligands like glitazones. With extended use, it is hoped that these effects will reduce the risk of long-term cardiovascular complications. PPAR alpha and gamma stimulation play complementary roles in the prevention of atherosclerosis. Cholesterol accumulation in macrophages located in the endothelium is a crucial step in the formation of atherosclerosis. PPAR gamma activation is necessary for the efflux of cholesterol from macrophage foam cells. Cholesterol taken up by HDL particles containing apolipoportein A-1 is transported to the liver to be disposed of as bile acids [3,15,17]. PPAR alpha agonists, on the other hand, speed up the transfer of cholesterol from macrophages to particles containing apolipoportein A-1 [3,16,20].
Thus, compounds with dual PPAR alpha/PPAR gamma activity appear well-suited for the treatment of diabetic patients with the additional risk factor of dyslipidemia. The finding that PPAR agonists play a role in regulating other processes, such as inflammation, vascular function, and vascular remodeling, has highlighted further potential indications for these agents [16,17]. So far, therefore, a relatively high number of dual PPAR alpha and PPAR gamma agonists have been described [3,21-25]. The new generation of dual-action PPARs – the glitazars which target PPAR-gamma and PPAR – alpha (muraglitazar and tesaglitazar) are on deck in late-stage clinical trials and may be effective in reducing cardiovascular risk, but their long-term clinical effects are still unknown. A number of glitazars have problems in late stage clinical trials because of serious side-effects (including ragaglitazar and farglitazar).
The bezafibrate lessons: feasibility of dual and pan-PPAR co-agonism in a clinical setting
The old and well known lipid-lowering fibric acid derivative bezafibrate is the first clinically tested pan – (alpha, beta/delta, gamma) PPAR activator [26-33]. It is a sole pan PPAR activator with more than a quarter of a century of a therapeutic experience with a good safety profile. Therefore, bezafibrate could be considered (indeed, as a "post hoc" understanding) as an "archetype" of a clinically tested pan-PPAR ligand. In patients with relevant metabolic abnormalities it is expected to improve both insulin sensitivity and the blood lipid profile and probably reduce the risk of long-term cardiovascular complications. In addition, we can expect prevention of overweight development due to its PPAR-beta/delta properties.
So, which are the data regarding bezafibrate administration? In a large trial in 1568 men with lower extremity arterial disease, bezafibrate reduced the severity of intermittent claudication for up to three years [34]. In general, the incidence of coronary heart disease in patients on bezafibrate has tended to be lower, but this tendency did not reach statistical significance. However, bezafibrate had significantly reduced the incidence of non-fatal coronary events, particularly in those aged <65 years at entry, in whom all coronary events may also be reduced [34]. In two other independent studies bezafibrate decreased the rate of progression of coronary atherosclerosis and decreased coronary events rate [35,36]. In the the Bezafibrate Infarction Prevention (BIP) study an overall trend of a 9.4% reduction of the incidence of primary end point (fatal or non-fatal myocardial infarction or sudden death) was observed. The reduction in the primary end point in 459 patients with high baseline triglycerides (200 mg/dL or more) was significant [37].
Our new data demonstrate that bezafibrate can significantly reduce the incidence of myocardial infarction (MI) in patients with metabolic syndrome [38]. The decrease in MI incidence among patients on bezafibrate was reflected in a trend to a late risk reduction of cardiac mortality during a long-term follow-up period. This tendency was strengthened in patients with augmented features (at least 4 risk factors for metabolic syndrome) of metabolic syndrome (56% reduction of cardiac mortality during 8-year follow-up). It is interesting that in patients without metabolic syndrome this favorable effect was not presented: There was no significant difference in the cardiovascular end-points between bezafibrate and placebo groups.
Previous observations have shown beneficial effects of bezafibrate on glucose and insulin metabolism [39-41]. Recently, we have shown thata pharmacological intervention with bezafibrate decreased the incidence and delayed the onset of type 2 diabetes in patients with impaired fasting glucose levels, and in obese patients over a long-term follow-up period [42,43]. In the BIP study the rates of adverse events were similar in both study groups [37]. Thus, bezafibrate treatment was safe in addition to being effective in diabetes prevention. Moreover, there was no significant change in mean body mass index values in either the bezafibrate or the placebo group during the follow-up [38,42,43].
Therefore, the pan – (alpha, beta, gamma) PPAR activator bezafibrate leads to a considerable raising of HDL cholesterol and a reduction of triglycerides, improves insulin sensitivity and reduces blood glucose level, significantly lowering the incidence of cardiovascular events and new diabetes in patients with features of metabolic syndrome over a long-term follow-up period. We conclude that clinical evidences obtained from bezafibrate-based studies strongly support the concept of a pan-PPAR therapeutic approach to conditions which comprise the metabolic syndrome. However, from a biochemical point of view, bezafibrate is PPAR ligand with a relatively low potency. We believe that more powerful compounds with pan-PPAR activity and proven long-term safety should be highly effective in a clinical setting of patients with coexisting relevant lipid and glucose metabolism disorders.
List of abbreviations used
BIP – Bezafibrate Infarction Prevention
HDL – high density lipoprotein
LDL – low density lipoprotein
MI – myocardial infarction
PPAR – peroxisome proliferator-activated receptor
RXR – retinoic acid receptor
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors have equally contributed in the conception and drafting of the manuscript.
Acknowledgements
This work was supported in part by the Cardiovascular Diabetology Research Foundation (RA 58-040-684-1), Holon, Israel, and the Research Authority of Tel-Aviv University (Dr. Ziternick and Haia Silva Ziternick Fund, grants 01250238 and 01250239).
==== Refs
Vamecq J Latruffe N Medical significance of peroxisome proliferator-activated receptors Lancet 1999 354 141 148 10408502 10.1016/S0140-6736(98)10364-1
Tenenbaum A Fisman EZ Motro M Metabolic syndrome and type 2 diabetes mellitus: focus on peroxisome proliferator activated receptors (PPAR) Cardiovasc Diabetol 2003 2 4 12834541 10.1186/1475-2840-2-4
Berger JP Akiyama TE Meinke PT PPARs: Therapeutic targets for metabolic disease Trends Pharmacol Sci 2005 26 244 251 15860371 10.1016/j.tips.2005.03.003
Kota BP Huang TH Roufogalis BD An overview on biological mechanisms of PPARs Pharmacol Res 2005 51 85 94 15629253 10.1016/j.phrs.2004.07.012
Fruchart JC Staels B Duriez P The role of fibric acids in atherosclerosis Curr Atheroscler Rep 2001 3 83 92 11123853
Desvergne B Michalik L Wahli W Be fit or be sick: peroxisome proliferator-activated receptors are down the road Mol Endocrinol 2004 18 1321 1332 15087471 10.1210/me.2004-0088
Argmann CA Cock TA Auwerx J Peroxisome proliferator-activated receptor gamma: the more the merrier? Eur J Clin Invest 2005 35 82 92 15667578 10.1111/j.1365-2362.2005.01456.x
Auwerx J PPARgamma, the ultimate thrifty gene Diabetologia 1999 42 1033 1049 10447513 10.1007/s001250051268
Lazar MA PPAR gamma, 10 years later Biochimie 2005 87 9 13 15733730 10.1016/j.biochi.2004.10.021
Fredenrich A Grimaldi PA PPAR delta: an uncompletely known nuclear receptor Diabetes Metab 2005 31 23 27 15803109
Burdick AD Kim DJ Peraza MA Gonzalez FJ Peters JM The role of peroxisome proliferator-activated receptor-beta/delta in epithelial cell growth and differentiation Cell Signal 2005, Aug 15 16109478
Staels B Fruchart JC Therapeutic roles of peroxisome proliferator-activated receptor agonists Diabetes 2005 54 2460 2470 16046315
Israelian-Konaraki Z Reaven PD Peroxisome proliferator-activated receptor-alpha and atherosclerosis: from basic mechanisms to clinical implications Cardiology 2005 103 1 9 15528894 10.1159/000081845
Forcheron F Cachefo A Thevenon S Pinteur C Beylot M Mechanisms of the triglyceride- and cholesterol-lowering effect of fenofibrate in hyperlipidemic type 2 diabetic patients Diabetes 2002 51 3486 3491 12453904
Chinetti G Lestavel S Fruchart JC Clavey V Staels B Peroxisome proliferator-activated receptor alpha reduces cholesterol esterification in macrophages Circ Res 2003 92 212 217 12574149 10.1161/01.RES.0000053386.46813.E9
Despres JP Lemieux I Robins SJ Role of fibric acid derivatives in the management of risk factors for coronary heart disease Drugs 2004 64 2177 2198 15456334
Giannini S Serio M Galli A Pleiotropic effects of thiazolidinediones: taking a look beyond antidiabetic activity J Endocrinol Invest 2004 27 982 991 15762051
Staels B PPARgamma and atherosclerosis Curr Med Res Opin 2005 21 S13 20 15811195 10.1185/030079905X36440
Luquet S Gaudel C Holst D Lopez-Soriano J Jehl-Pietri C Fredenrich A Grimaldi PA Roles of PPAR delta in lipid absorption and metabolism: a new target for the treatment of type 2 diabetes Biochim Biophys Acta 2005 1740 313 317 15949697
Ruan XZ Moorhead JF Fernando R Wheeler DC Powis SH Varghese Z PPAR agonists protect mesangial cells from interleukin 1beta-induced intracellular lipid accumulation by activating the ABCA1 cholesterol efflux pathway J Am Soc Nephrol 2003 14 593 600 12595494 10.1097/01.ASN.0000050414.52908.DA
Goldfarb B PPAR benefits beyond glucose control. Cardiovascular effects may be even greater with next generation DOC News 2005 2 14
Bays H Stein EA Pharmacotherapy for dyslipidaemia – current therapies and future agents Expert Opin Pharmacother 2003 4 1901 1938 14596646
Bailey CJ New pharmacologic agents for diabetes Curr Diab Rep 2001 1 119 126 12643107
Pegorier JP [PPAR receptors and insulin sensitivity: new agonists in development] Ann Endocrinol (Paris) 2005 66 1S10 17 15959400
Darves B Muraglitazar may help lower glucose and cholesterol levels in type 2 diabetes Medscape Medical News, ADA 65th Annual Scientific Sessions: Abstracts 967, 14-OR 2005
Peters JM Aoyama T Burns AM Gonzalez FJ Bezafibrate is a dual ligand for PPARalpha and PPARbeta: studies using null mice Biochim Biophys Acta 2003 1632 80 89 12782154
Cabrero A Alegret M Sanchez RM Adzet T Laguna JC Vazquez M Bezafibrate reduces mRNA levels of adipocyte markers and increases fatty acid oxidation in primary culture of adipocytes Diabetes 2001 50 1883 1890 11473052
Poirier H Niot I Monnot MC Braissant O Meunier-Durmort C Costet P Pineau T Wahli W Willson TM Besnard P Differential involvement of peroxisome-proliferator-activated receptors alpha and delta in fibrate and fatty-acid-mediated inductions of the gene encoding liver fatty-acid-binding protein in the liver and the small intestine Biochem J 2001 355 481 488 11284737 10.1042/0264-6021:3550481
Vazquez M Roglans N Cabrero A Rodriguez C Adzet T Alegret M Sanchez RM Laguna JC Bezafibrate induces acyl-CoA oxidase mRNA levels and fatty acid peroxisomal beta-oxidation in rat white adipose tissue Mol Cell Biochem 2001 216 71 78 11216866 10.1023/A:1011060615234
Krey G Braissant O L'Horset F Kalkhoven E Perroud M Parker MG Wahli W Fatty acids, eicosanoids, and hypolipidemic agents identified as ligands of peroxisome proliferator-activated receptors by coactivator-dependent receptor ligand assay Mol Endocrinol 1997 11 779 791 9171241 10.1210/me.11.6.779
Moya-Camarena SY Van den Heuvel JP Belury MA Conjugated linoleic acid activates peroxisome proliferator-activated receptor alpha and beta subtypes but does not induce hepatic peroxisome proliferation in Sprague-Dawley rats Biochim Biophys Acta 1999 1436 331 342 9989264
Willson TM Brown PJ Sternbach DD Henke BR The PPARs: from orphan receptors to drug discovery J Med Chem 2000 43 527 550 10691680 10.1021/jm990554g
Berger J Moller DE The mechanisms of action of PPARs Annu Rev Med 2002 53 409 435 11818483 10.1146/annurev.med.53.082901.104018
Meade T Zuhrie R Cook C Cooper J Bezafibrate in men with lower extremity arterial disease: randomised controlled trial BMJ 2002 325 1139 12433762 10.1136/bmj.325.7373.1139
Ericsson CG Nilsson J Grip L Svane B Hamsten A Effect of bezafibrate treatment over five years on coronary plaques causing 20% to 50% diameter narrowing (The Bezafibrate Coronary Atherosclerosis Intervention Trial [BECAIT]) Am J Cardiol 1997 8 1125 1129 9359536 10.1016/S0002-9149(97)00626-7
Elkeles RS Diamond JR Poulter C Dhanjil S Nicolaides AN Mahmood S Richmond W Mather H Sharp P Feher MD Cardiovascular outcomes in type 2 diabetes. A double-blind placebo-controlled study of bezafibrate: the St. Mary's, Ealing, Northwick Park Diabetes Cardiovascular Disease Prevention (SENDCAP) Study Diabetes Care 1998 21 641 648 9571357
Secondary prevention by raising HDL cholesterol and reducing triglycerides in patients with coronary artery disease: the Bezafibrate Infarction Prevention (BIP) study Circulation 2000 102 21 27 10880410
Tenenbaum A Motro M Fisman EZ Tanne D Boyko V Behar S Bezafibrate for the secondary prevention of myocardial infarction in patients with metabolic syndrome Arch Intern Med 2005 165 1154 1160 15911729 10.1001/archinte.165.10.1154
Taniguchi A Fukushima M Sakai M Tokuyama K Nagata I Fukunaga A Kishimoto H Doi K Yamashita Y Matsuura T Kitatani N Okumura T Nagasaka S Nakaishi S Nakai Y Effects of bezafibrate on insulin sensitivity and insulin secretion in non-obese Japanese type 2 diabetic patients Metabolism 2001 50 477 480 11288046 10.1053/meta.2001.21028
Jonkers IJ Mohrschladt MF Westendorp RG van der Laarse A Smelt AH Severe hypertriglyceridemia with insulin resistance is associated with systemic inflammation: reversal with bezafibrate therapy in a randomized controlled trial Am J Med 2002 112 275 280 11893366 10.1016/S0002-9343(01)01123-8
Kim JI Tsujino T Fujioka Y Saito K Yokoyama M Bezafibrate improves hypertension and insulin sensitivity in humans Hypertens Res 2003 26 307 313 12733699 10.1291/hypres.26.307
Tenenbaum A Motro M Fisman EZ Schwammenthal E Adler Y Goldenberg I Leor J Boyko V Mandelzweig L Behar S Peroxisome proliferator-activated receptors ligand bezafibrate for prevention of type 2 diabetes mellitus in patients with coronary artery disease Circulation 2004 109 2197 2202 15123532 10.1161/01.CIR.0000126824.12785.B6
Tenenbaum A Motro M Fisman EZ Adler Y Shemesh J Tanne D Leor J Boyko V Schwammenthal E Behar S Effect of bezafibrate on incidence of type 2 diabetes mellitus in obese patients Eur Heart J 2005 26 2032 2038 15872029
|
16168052
|
PMC1236941
|
CC BY
|
2021-01-04 16:26:15
|
no
|
Cardiovasc Diabetol. 2005 Sep 16; 4:14
|
utf-8
|
Cardiovasc Diabetol
| 2,005 |
10.1186/1475-2840-4-14
|
oa_comm
|
==== Front
Cardiovasc UltrasoundCardiovascular Ultrasound1476-7120BioMed Central London 1476-7120-3-241613733210.1186/1476-7120-3-24ResearchWhat parameters affect left ventricular diastolic flow propagation velocity? in vitro studies using color m-mode doppler echocardiography Ogawa Toshihiro [email protected] Lawrence N [email protected] David K [email protected] Ajit P [email protected] Renee L [email protected] Cheryl K [email protected] Julius M [email protected] St. John Hospital & Medical Center, Detroit, MI, USA2 Vivitro Systems, Inc., Victoria, BC, Canada3 Georgia Institute of Technology, Atlanta, GA, USA2005 1 9 2005 3 24 24 20 4 2005 1 9 2005 Copyright © 2005 Ogawa et al; licensee BioMed Central Ltd.2005Ogawa 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
Insufficient data describe the relationship of hemodynamic parameters to left ventricular (LV) diastolic flow propagation velocity (Vp) measured using color M-mode Doppler echocardiography.
Methods
An in vitro LV model used to simulate LV diastolic inflow with Vp measured under conditions of varying: 1) Stroke volume, 2) heart rate (HR), 3) LV volume, 4) LV compliance, and 5) transmitral flow (TMF) waveforms (Type 1: constant low diastasis flow and Type 2: no diastasis flow).
Results
Univariate analysis revealed excellent correlations of Vp with stroke volume (r = 0.98), LV compliance (r = 0.94), and HR with Type 1 TMF (r = 0.97). However, with Type 2 TMF, HR was not associated with Vp. LV volume was not related to Vp under low compliance, but inversely related to Vp under high compliance conditions (r = -0.56).
Conclusion
These in vitro findings may help elucidate the relationship of hemodynamic parameters to early diastolic LV filling.
==== Body
Background
The flow Vp of early diastolic LV inflow measured using color M-mode Doppler echocardiography (CMD) has been recognized as a useful measure of LV relaxation. [1-4] Vp by CMD has also been reported to correlate with the time constant of isovolumic relaxation (τ). [1,3,4] However, there are few published data describing the relationship of hemodynamic parameters to Vp. In the clinical and in vivo experimental situations, it is difficult to evaluate the influence of any single hemodynamic variable on Vp because changing one variable often is associated with changes in other variables, such as LV contraction and heart rate. Therefore, to better understand the influence of changes in various hemodynamic parameters on Vp, we performed in vitro CMD studies using a mechanical LV model which allowed us to produce isolated changes in the hemodynamic parameters.
Methods
Model
We used a customized commercially-available LV model (Superdup'r SD1002, Vivitro Systems Inc., Victoria, BC, Canada), which was modified to facilitate ultrasound interrogation by placing echo-transducer ports at the side of the LV apex and left atrium. The LV diaphragm, made of silicone rubber, had a hemi-ellipsoid shape (long-axis dimension = 53 mm, base diameter = 44 mm, thickness = 0.6 mm). The normal baseline LV volume contained 150 ml of saline. The hydraulic chamber surrounding the LV diaphragm was filled with distilled water. The LV diaphragm was contracted and expanded by a computer-programmed piston pump which controlled the volume of the hydraulic fluid. Pericardial bileaflet bio-prosthetic valves (diameter = 21 mm) were mounted at the aortic and mitral valve sites. Aortic, LV, and left atrial pressure, and aortic or mitral flow rate were monitored during experiments.
Waveform programming
Aortic and mitral waveforms were programmed to mimic various physiological conditions. We programmed two different types of transmitral waveforms: Type 1 (with constant low velocity diastasis flow) and Type 2 (no mitral flow during the diastasis period; not shown in Figure 1) (Figure 1). Although HR was varied, transmitral waveforms were programmed to keep E (early diastolic) and A (late diastolic, or atrial) wave shape, acceleration time, and deceleration time constant at the mitral valve site. Aortic flow durations were set at 35% of cycle length at HR = 40 and 50/min, 27% at 60/min, 31% at 70/min, and 36% at 80/min.
Figure 1 Type 1 transmitral flow velocity diagram: Constant diastasis flow. Heart rate = 60/min.
CMD measurements
CMD measurements were performed using a Vivid ultrasound unit (General Electric Vingmed Ultrasound, Horten, Norway). The transducer (2.5 MHz) was positioned at the LV apex site. The distance between transducer and mitral valve was 7.6 cm. The CMD red-blue interface (aliasing) velocity was set at 92 cm/sec to avoid signal "bleeding". CMD sweep speed was 200 mm/sec. The echocardiography machine settings were kept constant during all experiments. All CMD images were recorded on an optical disc or CD-ROM, and off-line analysis was performed using commercial software (EchoPac for Vivid 7). The Vp was measured as the slope of the first aliasing velocity from the mitral annulus to 4 cm distal in the left ventricle in early diastole. Vp measurements were performed 10 times for each flow condition.
Hemodynamic variables
The following ranges of hemodynamic variables were studied: 1) stroke volume (SV): 50, 60, 70, 80, and 90 ml; 2) heart rate (HR): 40, 50, 60, 70, and 80/min; 3) baseline LV volume (LVV): 130, 150, 180, 200, and 200 ml; 4) LV compliance: five conditions; and 5) transmitral flow (TMF) waveforms (Type 1: featuring constant low diastasis flow between early and late diastolic waves and Type 2: featuring no diastasis flow). LVV was changed by adding distilled water to, or sucking hydraulic liquid from, the hydraulic chamber. Adding 0, 10, 20, 30, or 40 ml of air into the hydraulic chamber and sucking the same volume of hydraulic liquid from the hydraulic chamber changed LV compliance – specifically, the greater the volume of air, the greater the LV compliance. Baseline conditions were: Type 1 waveform, SV = 70 ml, HR = 60/min, LVV = 150 ml, and an LV compliance with 0 ml of air in the hydraulic chamber. LV compliance was calculated as volume change divided by pressure change during the period from the LV-left atrial pressure crossover-point to the minimum LV pressure in early diastole. The LV compliance (ml/mmHg) under 5 experimental conditions was calculated as follows: 0 ml air, 8.32 × 10-3; 10 ml air, 9.75 × 10-3; 20 ml air, 16.34 × 10-3; 30 ml air, 17.05 × 10-3; and 40 ml air, 19.92 × 10-3.
Measurement Variability
Vp measurements were made using commercially available software. Vp was measured on two separate occasions on 26 images stored on disc by one reader to estimate intra-reader measurement variability. A second observer, blinded to the measurements made by the first reader, also measured Vp on these 26 images to calculate inter-reader variability.
Statistical analysis
The association of hemodynamic parameters with Vp was analyzed using Student's paired t test and univariate regression analysis. All calculated P values were two-tailed, and a value of P < 0.05 was considered to indicate statistical significance.
Results
Measurement variability
The mean difference (+ or - SD) of Vp measurements for the same reader was 1.56 ± 0.88 cm/sec (2.0 ± 1.1% mean intra-reader variability). The mean difference (± SD) of Vp measurements between two readers was 1.79 ± 1.16 cm/sec (2.2 ± 1.4% mean inter-reader measurement variability).
Relationship between hemodynamic variables and Vp
SV (r = 0.98, p < 0.0001) and LV compliance (r = 0.95, p < 0.0001) both showed excellent correlations with Vp (Figures 2 and 3). With Type 1 transmitral waveform conditions, Vp increased as HR increased (r = 0.97, p < 0.0001) (Figure 4). However, with Type 2 waveform conditions, HR was not associated with Vp (r = 0.15, p = NS) (Figure 5). The LV volume was not associated with Vp at SV 50 ml (r = -0.08, p = NS) and SV 70 ml (r = 0.04, p = NS) under low compliance conditions (no added air, Figures 6 and 7). However, under high compliance conditions (40 ml air in the hydraulic chamber), LV volume was inversely associated with Vp at SV 50 ml (r = 0.59, p < 0.0001) and SV 70 ml (r = -0.56, p < 0.0001, Figures 8 and 9).
Figure 2 This graph shows a strong relationship between stroke volume (SV) and flow propagation velocity (Vp) (r = 0.98, p < 0.0001).
Figure 3 This graph shows an excellent correlation between LV compliance and Vp (r = 0.95, p < 0.0001).
Figure 4 Relationship between heart rate (HR) and Vp with a Type 1 transmitral flow pattern: There was a good correlation between HR and Vp (r = 0.97, p < 0.0001).
Figure 5 Relationship between heart rate (HR) and Vp with a Type 2 transmitral flow pattern: HR was not associated with Vp (r = 0.15, p = NS).
Figure 6 Relationship between LV volume (LVV) and Vp: Under low compliance conditions (no added air) and 70 ml stroke volume, LVV was not associated with Vp (r = 0.04, p = NS).
Figure 7 Relationship between LVV and Vp: Under low compliance conditions and 50 ml stroke volume, LVV was not associated with Vp (r = -0.08, p = NS).
Figure 8 Relationship between LVV and Vp: Under high compliance conditions (40 ml air in hydraulic chamber) and SV = 50 ml, LVV was inversely associated with Vp (r = -0.59, p < 0.0001).
Figure 9 Relationship between LVV and Vp: Under high compliance conditions and SV = 70 ml, LVV was inversely associated with Vp (r = -0.56, p < 0.0001).
Discussion
Stroke Volume and Vp
Many investigators have studied diastolic suction. They have characterized diastolic suction as ventricular filling under zero source pressure for filling or peak diastolic negative pressure. The elastic recoil of the relaxing ventricular wall is considered to be the source of the ventricular suction force [5,6]. Courtois reported that the intraventricular pressure gradient (IVPG) between the apex and base of the left ventricle during early diastolic filling was reduced in myocardial ischemia cases; they speculated that IVPGs were related to elastic recoil [7]. Greenberg reported that IVPG was correlated with flow propagation velocity [8]. Smiseth demonstrated that IVPG also correlated with peak early transmitral flow and stroke volume in canine studies [9]. More recently, in their clinical investigation, Ohte showed a good relationship between Vp and LV end-systolic volume or LV ejection fraction [10]. The positive relationship between Vp and stroke volume documented in the current study is consistent with these reports.
LV compliance and Vp
Using computer simulations, Lemmon showed that the effect of delayed relaxation was to decrease the early filling propagation, and this decrease was larger when the stiffness of the ventricle was increased [11]. Vierendeels also demonstrated that higher LV stiffness was associated with a smaller Vp in their computer simulations [12,13]. Our findings are consistent with these previous studies.
Heart rate and early diastolic LV filling
The difference in HR relationship to Vp with our two transmitral waveform models (Type 1 and Type 2) may be due to the fact that the waveform program software automatically modifies transmitral velocity-time integral to be equal to the aortic velocity-time integral. There is no in vivo study confirming a HR effect on Vp. However, some previous clinical studies and human or canine studies employing atrial pacing demonstrated that increasing HR was not associated with a change in mitral peak E velocity on pulsed wave Doppler [14-18]. Cheng reported that in their canine studies using atrial pacing, as HR increased from 100 to 160/min, left atrial pressure dropped in a monotonic fashion and τ also decreased [19]. Therefore, relaxation filling was unchanged at HR up to 160/min, consistent with the opposite and counterbalancing influences of HR and τ [19]. In our studies, left atrial pressure was constant though HR changed. However, in an epidemiologic study [20] and in an atrial pacing study of patients with DDD pacemakers [21], the transmitral peak E velocity decreased as HR increased. Many factors affect the transmitral flow profile, and this discrepancy may be based on other factors (e.g., a normal vs. pathological heart, myocardial stiffness, preload, and LV geometry).
LV volume and Vp
Steen reported, in a mechanical LV model similar to ours, that a larger ventricle had a higher Vp than did a smaller ventricle for the same transmitral peak E wave velocity [22]. The difference between Steen's results and ours may be due to: 1) the presence of an LV outflow tract in our LV model, and 2) the fact that our model included a flow circuit, but their simple model did not include serial LV inflow, which would have mitigated any adverse effect of inertia on the LV inflow profile. In their finite element model, Sunagawa reported that the combination of increased LV mass, low stiffness and low strain axis parameter, was associated with increased SV with ventricular volume reduction [23]. Our results related to the LVV and Vp, are consistent with the latter study.
Study limitations
Our hydraulic LV model had various limitations in simulating the human left ventricle. Specifically, the LV had negligible mass, no papillary muscles and, higher volume compared with the normal human left ventricle, and lacked the LV twisting motion and left atrial contraction with changes in left atrial volume. Also, during LV filling, the position and size of the mitral valve annulus were fixed in our model.
Conclusion
We used a mechanical LV model to study how Vp measured by color M-mode Doppler echocardiography varied with changes in stroke volume, heart rate, LV volume, LV compliance, and mitral waveform. Our results documented important relationships between various physiologic parameters and Vp. These studies of Vp should help to elucidate the physiology of early diastolic LV filling. Further studies are needed to investigate the relationship between factors such as left atrial pressure, severity of mitral regurgitation, and mitral annulus size and Vp.
Declaration of Competing Interests
Drs. Toshihiro Ogawa, Ajit Yoganathan, Cheryl Nordstrom, Julius Gardin, and Renee Bess all declare that they have no competing interests. Dr. David Walker and Mr. Lawrence Scotten state that during the time of the performance of these studies, they were employees of Vivitro Systems, Inc., which was the manufacturer of the in vitro flow model (Superdup'r SD1002) used in this study.
==== Refs
Brun P Tribouilloy C Duval AM Iseriu L Megurira A Pelle G Dubois-Rande JL Left ventricular flow propagation during early filling is related to wall relaxation: a color M-mode Doppler analysis J Am Coll Cardiol 1992 20 420 432 1634681
Stugaard M Risoe C Halfdan I Smiseth OA Intraventricular early diastolic filling during acute myocardial ischemia: assessment by multigated color M-mode Doppler echocardiography Circulation 1993 88 2705 2713 8252682
Takatsuji H Mikami T Urasawa K Teranishi J-I Onozuka H Takagi C Makita Y Matsuo H Kusuoka H Kitabatake A A new approach for evaluation of left ventricular diastolic function: Spatial and temporal analysis of left ventricular filling flow propagation by color M-mode Doppler echocardiography J Am Coll Cardiol 1996 27 365 371 8557907 10.1016/0735-1097(96)81240-X
Garcia MJ Smedira NG Greenberg NL Main M Firstenberg MS Odabashian J Thomas JD Color M-mode Doppler flow propagation velocity is a preload insensitive index of left ventricular relaxation: animal and human validation J Am Coll Cardiol 2000 35 201 208 10636281 10.1016/S0735-1097(99)00503-3
Yellin EL Hori M Yorcan C Sonnenblick EH Gabbay S Frater RW Left ventricular relaxation in the filling and nonfilling intact canine heart Am J Physiol 1986 250 H620 629 3963218
Suga H Goto Y Igarashi Y Yamada O Nozawa T Yasumura Y Ventricular suction under zero pressure for filling Am J Physiol 1986 251 H47 55 3728698
Courtois M Kovacs SJ JrLudbrook PA Physiological early diastolic intraventricular pressure gradient is lost during acute myocardial ischemia Circulation 1990 81 1688 1696 2331773
Greenberg NL Vandervoort PM Thomas JD Noninvasive assessment of diastolic intraventricular pressure gradients using color Doppler M-mode echocardiograph Computers in cardiology 1995 Los Angeles, CA: IEE Computer Society Press 465 468
Smiseth OA Steine K Sandbaek G Stugaard M Gjolberg TO Mechanics of intraventricular filling: study of LV early diastolic pressure gradients and flow velocities Am J Physiol 1998 275 H1062 1069 9724314
Ohte N Naria H Akita S Kurokawa K Hayano J Kimura G Striking effect of left ventricular systolic performance on propagation velocity of left ventricular early diastolic filling flow J Am Soc Echocardiogr 2001 14 1070 1074 11696830
Lemmon JD Yoganathan AP Computational modeling of left heart diastolic function: examination of ventricular dysfunction J Biomech Eng 2000 122 297 303 11036551 10.1115/1.1286559
Vierendeels JA Riemslagh K Dick E Computer simulation of intraventricular flow and pressure gradients during diastole J Biomech Eng 2000 122 667 674 11192390 10.1115/1.1318941
Vierendeels JA Dick E Verdonck PR Hydrodynamics of color M-mode Doppler flow wave propagation velocity V (p): a computer study J Am Soc Echocardiogr 2002 15 219 224 11875384 10.1067/mje.2002.115456
Harrison MR Clifton GD Pennel AT DeMaria AN Effect of heart rate on left ventricular diastolic transmitral flow velocity patterns assessed by Doppler echocardiography in normal subjects Am J Cardiol 1991 67 622 627 2000796 10.1016/0002-9149(91)90902-W
Appleton CP Influence of incremental changes in heart rate on mitral flow velocity: assessment in lightly sedated, conscious dogs J Am Coll Cardiol 1991 17 227 236 1987230
Oniki T Hashimoto Y Shimizu S Kahuta T Yajima M Numano F Effect of increasing heart rate on Doppler indices of left ventricular performance in healthy men Br Heart J 1992 68 425 429 1449930
Yu Ch M Sanderson JE Right and left ventricular diastolic function in patients with and without heart failure: effect of age, sec, heart rate, and respiration on Doppler-derived measurements Am Heart J 1997 134 426 434 9327698
Schober KE Fuentes VL Effects of age, body weight, and heart rate on transmitral and pulmonary venous flow in clinically normal dogs Am J Vet Res 2001 62 1447 1454 11560276
Cheng CP Freeman GL Santamore WP Constantinescu MS Little WC Effect of loading conditions, contractile state, and heart rate on early diastolic left ventricular filling in conscious dogs Circ Res 1990 66 814 823 1689618
Galderisi M Benjamin EJ Evans JC D'Agostino RB Fuller DL Lehman B Levy D Impact of heart rate and PR interval on Doppler indexes of left ventricular diastolic filling in an elderly cohort (the Framingham heart study) Am J Cardiol 1993 72 1183 1187 8237811 10.1016/0002-9149(93)90991-K
Gillam LD Homma S Novick SS Rediker DE Eagle KA. The influence of heart rate on Doppler mitral inflow patterns Circulation 1987 76 IV 123
Steen T Steen S Filling of a model left ventricle studied by colour M-Mode Doppler Cardiovasc Res 1994 28 1821 1827 7867035 7867035
Sugimachi M Sunagawa K Effects of partial left ventriculectomy on left ventricular pump function studied by theoretical analysis J Card Surg 2001 16 24 29 11713853
|
16137332
|
PMC1236942
|
CC BY
|
2021-01-04 16:38:31
|
no
|
Cardiovasc Ultrasound. 2005 Sep 1; 3:24
|
utf-8
|
Cardiovasc Ultrasound
| 2,005 |
10.1186/1476-7120-3-24
|
oa_comm
|
==== Front
Cardiovasc UltrasoundCardiovascular Ultrasound1476-7120BioMed Central London 1476-7120-3-241613733210.1186/1476-7120-3-24ResearchWhat parameters affect left ventricular diastolic flow propagation velocity? in vitro studies using color m-mode doppler echocardiography Ogawa Toshihiro [email protected] Lawrence N [email protected] David K [email protected] Ajit P [email protected] Renee L [email protected] Cheryl K [email protected] Julius M [email protected] St. John Hospital & Medical Center, Detroit, MI, USA2 Vivitro Systems, Inc., Victoria, BC, Canada3 Georgia Institute of Technology, Atlanta, GA, USA2005 1 9 2005 3 24 24 20 4 2005 1 9 2005 Copyright © 2005 Ogawa et al; licensee BioMed Central Ltd.2005Ogawa 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
Insufficient data describe the relationship of hemodynamic parameters to left ventricular (LV) diastolic flow propagation velocity (Vp) measured using color M-mode Doppler echocardiography.
Methods
An in vitro LV model used to simulate LV diastolic inflow with Vp measured under conditions of varying: 1) Stroke volume, 2) heart rate (HR), 3) LV volume, 4) LV compliance, and 5) transmitral flow (TMF) waveforms (Type 1: constant low diastasis flow and Type 2: no diastasis flow).
Results
Univariate analysis revealed excellent correlations of Vp with stroke volume (r = 0.98), LV compliance (r = 0.94), and HR with Type 1 TMF (r = 0.97). However, with Type 2 TMF, HR was not associated with Vp. LV volume was not related to Vp under low compliance, but inversely related to Vp under high compliance conditions (r = -0.56).
Conclusion
These in vitro findings may help elucidate the relationship of hemodynamic parameters to early diastolic LV filling.
==== Body
Background
The flow Vp of early diastolic LV inflow measured using color M-mode Doppler echocardiography (CMD) has been recognized as a useful measure of LV relaxation. [1-4] Vp by CMD has also been reported to correlate with the time constant of isovolumic relaxation (τ). [1,3,4] However, there are few published data describing the relationship of hemodynamic parameters to Vp. In the clinical and in vivo experimental situations, it is difficult to evaluate the influence of any single hemodynamic variable on Vp because changing one variable often is associated with changes in other variables, such as LV contraction and heart rate. Therefore, to better understand the influence of changes in various hemodynamic parameters on Vp, we performed in vitro CMD studies using a mechanical LV model which allowed us to produce isolated changes in the hemodynamic parameters.
Methods
Model
We used a customized commercially-available LV model (Superdup'r SD1002, Vivitro Systems Inc., Victoria, BC, Canada), which was modified to facilitate ultrasound interrogation by placing echo-transducer ports at the side of the LV apex and left atrium. The LV diaphragm, made of silicone rubber, had a hemi-ellipsoid shape (long-axis dimension = 53 mm, base diameter = 44 mm, thickness = 0.6 mm). The normal baseline LV volume contained 150 ml of saline. The hydraulic chamber surrounding the LV diaphragm was filled with distilled water. The LV diaphragm was contracted and expanded by a computer-programmed piston pump which controlled the volume of the hydraulic fluid. Pericardial bileaflet bio-prosthetic valves (diameter = 21 mm) were mounted at the aortic and mitral valve sites. Aortic, LV, and left atrial pressure, and aortic or mitral flow rate were monitored during experiments.
Waveform programming
Aortic and mitral waveforms were programmed to mimic various physiological conditions. We programmed two different types of transmitral waveforms: Type 1 (with constant low velocity diastasis flow) and Type 2 (no mitral flow during the diastasis period; not shown in Figure 1) (Figure 1). Although HR was varied, transmitral waveforms were programmed to keep E (early diastolic) and A (late diastolic, or atrial) wave shape, acceleration time, and deceleration time constant at the mitral valve site. Aortic flow durations were set at 35% of cycle length at HR = 40 and 50/min, 27% at 60/min, 31% at 70/min, and 36% at 80/min.
Figure 1 Type 1 transmitral flow velocity diagram: Constant diastasis flow. Heart rate = 60/min.
CMD measurements
CMD measurements were performed using a Vivid ultrasound unit (General Electric Vingmed Ultrasound, Horten, Norway). The transducer (2.5 MHz) was positioned at the LV apex site. The distance between transducer and mitral valve was 7.6 cm. The CMD red-blue interface (aliasing) velocity was set at 92 cm/sec to avoid signal "bleeding". CMD sweep speed was 200 mm/sec. The echocardiography machine settings were kept constant during all experiments. All CMD images were recorded on an optical disc or CD-ROM, and off-line analysis was performed using commercial software (EchoPac for Vivid 7). The Vp was measured as the slope of the first aliasing velocity from the mitral annulus to 4 cm distal in the left ventricle in early diastole. Vp measurements were performed 10 times for each flow condition.
Hemodynamic variables
The following ranges of hemodynamic variables were studied: 1) stroke volume (SV): 50, 60, 70, 80, and 90 ml; 2) heart rate (HR): 40, 50, 60, 70, and 80/min; 3) baseline LV volume (LVV): 130, 150, 180, 200, and 200 ml; 4) LV compliance: five conditions; and 5) transmitral flow (TMF) waveforms (Type 1: featuring constant low diastasis flow between early and late diastolic waves and Type 2: featuring no diastasis flow). LVV was changed by adding distilled water to, or sucking hydraulic liquid from, the hydraulic chamber. Adding 0, 10, 20, 30, or 40 ml of air into the hydraulic chamber and sucking the same volume of hydraulic liquid from the hydraulic chamber changed LV compliance – specifically, the greater the volume of air, the greater the LV compliance. Baseline conditions were: Type 1 waveform, SV = 70 ml, HR = 60/min, LVV = 150 ml, and an LV compliance with 0 ml of air in the hydraulic chamber. LV compliance was calculated as volume change divided by pressure change during the period from the LV-left atrial pressure crossover-point to the minimum LV pressure in early diastole. The LV compliance (ml/mmHg) under 5 experimental conditions was calculated as follows: 0 ml air, 8.32 × 10-3; 10 ml air, 9.75 × 10-3; 20 ml air, 16.34 × 10-3; 30 ml air, 17.05 × 10-3; and 40 ml air, 19.92 × 10-3.
Measurement Variability
Vp measurements were made using commercially available software. Vp was measured on two separate occasions on 26 images stored on disc by one reader to estimate intra-reader measurement variability. A second observer, blinded to the measurements made by the first reader, also measured Vp on these 26 images to calculate inter-reader variability.
Statistical analysis
The association of hemodynamic parameters with Vp was analyzed using Student's paired t test and univariate regression analysis. All calculated P values were two-tailed, and a value of P < 0.05 was considered to indicate statistical significance.
Results
Measurement variability
The mean difference (+ or - SD) of Vp measurements for the same reader was 1.56 ± 0.88 cm/sec (2.0 ± 1.1% mean intra-reader variability). The mean difference (± SD) of Vp measurements between two readers was 1.79 ± 1.16 cm/sec (2.2 ± 1.4% mean inter-reader measurement variability).
Relationship between hemodynamic variables and Vp
SV (r = 0.98, p < 0.0001) and LV compliance (r = 0.95, p < 0.0001) both showed excellent correlations with Vp (Figures 2 and 3). With Type 1 transmitral waveform conditions, Vp increased as HR increased (r = 0.97, p < 0.0001) (Figure 4). However, with Type 2 waveform conditions, HR was not associated with Vp (r = 0.15, p = NS) (Figure 5). The LV volume was not associated with Vp at SV 50 ml (r = -0.08, p = NS) and SV 70 ml (r = 0.04, p = NS) under low compliance conditions (no added air, Figures 6 and 7). However, under high compliance conditions (40 ml air in the hydraulic chamber), LV volume was inversely associated with Vp at SV 50 ml (r = 0.59, p < 0.0001) and SV 70 ml (r = -0.56, p < 0.0001, Figures 8 and 9).
Figure 2 This graph shows a strong relationship between stroke volume (SV) and flow propagation velocity (Vp) (r = 0.98, p < 0.0001).
Figure 3 This graph shows an excellent correlation between LV compliance and Vp (r = 0.95, p < 0.0001).
Figure 4 Relationship between heart rate (HR) and Vp with a Type 1 transmitral flow pattern: There was a good correlation between HR and Vp (r = 0.97, p < 0.0001).
Figure 5 Relationship between heart rate (HR) and Vp with a Type 2 transmitral flow pattern: HR was not associated with Vp (r = 0.15, p = NS).
Figure 6 Relationship between LV volume (LVV) and Vp: Under low compliance conditions (no added air) and 70 ml stroke volume, LVV was not associated with Vp (r = 0.04, p = NS).
Figure 7 Relationship between LVV and Vp: Under low compliance conditions and 50 ml stroke volume, LVV was not associated with Vp (r = -0.08, p = NS).
Figure 8 Relationship between LVV and Vp: Under high compliance conditions (40 ml air in hydraulic chamber) and SV = 50 ml, LVV was inversely associated with Vp (r = -0.59, p < 0.0001).
Figure 9 Relationship between LVV and Vp: Under high compliance conditions and SV = 70 ml, LVV was inversely associated with Vp (r = -0.56, p < 0.0001).
Discussion
Stroke Volume and Vp
Many investigators have studied diastolic suction. They have characterized diastolic suction as ventricular filling under zero source pressure for filling or peak diastolic negative pressure. The elastic recoil of the relaxing ventricular wall is considered to be the source of the ventricular suction force [5,6]. Courtois reported that the intraventricular pressure gradient (IVPG) between the apex and base of the left ventricle during early diastolic filling was reduced in myocardial ischemia cases; they speculated that IVPGs were related to elastic recoil [7]. Greenberg reported that IVPG was correlated with flow propagation velocity [8]. Smiseth demonstrated that IVPG also correlated with peak early transmitral flow and stroke volume in canine studies [9]. More recently, in their clinical investigation, Ohte showed a good relationship between Vp and LV end-systolic volume or LV ejection fraction [10]. The positive relationship between Vp and stroke volume documented in the current study is consistent with these reports.
LV compliance and Vp
Using computer simulations, Lemmon showed that the effect of delayed relaxation was to decrease the early filling propagation, and this decrease was larger when the stiffness of the ventricle was increased [11]. Vierendeels also demonstrated that higher LV stiffness was associated with a smaller Vp in their computer simulations [12,13]. Our findings are consistent with these previous studies.
Heart rate and early diastolic LV filling
The difference in HR relationship to Vp with our two transmitral waveform models (Type 1 and Type 2) may be due to the fact that the waveform program software automatically modifies transmitral velocity-time integral to be equal to the aortic velocity-time integral. There is no in vivo study confirming a HR effect on Vp. However, some previous clinical studies and human or canine studies employing atrial pacing demonstrated that increasing HR was not associated with a change in mitral peak E velocity on pulsed wave Doppler [14-18]. Cheng reported that in their canine studies using atrial pacing, as HR increased from 100 to 160/min, left atrial pressure dropped in a monotonic fashion and τ also decreased [19]. Therefore, relaxation filling was unchanged at HR up to 160/min, consistent with the opposite and counterbalancing influences of HR and τ [19]. In our studies, left atrial pressure was constant though HR changed. However, in an epidemiologic study [20] and in an atrial pacing study of patients with DDD pacemakers [21], the transmitral peak E velocity decreased as HR increased. Many factors affect the transmitral flow profile, and this discrepancy may be based on other factors (e.g., a normal vs. pathological heart, myocardial stiffness, preload, and LV geometry).
LV volume and Vp
Steen reported, in a mechanical LV model similar to ours, that a larger ventricle had a higher Vp than did a smaller ventricle for the same transmitral peak E wave velocity [22]. The difference between Steen's results and ours may be due to: 1) the presence of an LV outflow tract in our LV model, and 2) the fact that our model included a flow circuit, but their simple model did not include serial LV inflow, which would have mitigated any adverse effect of inertia on the LV inflow profile. In their finite element model, Sunagawa reported that the combination of increased LV mass, low stiffness and low strain axis parameter, was associated with increased SV with ventricular volume reduction [23]. Our results related to the LVV and Vp, are consistent with the latter study.
Study limitations
Our hydraulic LV model had various limitations in simulating the human left ventricle. Specifically, the LV had negligible mass, no papillary muscles and, higher volume compared with the normal human left ventricle, and lacked the LV twisting motion and left atrial contraction with changes in left atrial volume. Also, during LV filling, the position and size of the mitral valve annulus were fixed in our model.
Conclusion
We used a mechanical LV model to study how Vp measured by color M-mode Doppler echocardiography varied with changes in stroke volume, heart rate, LV volume, LV compliance, and mitral waveform. Our results documented important relationships between various physiologic parameters and Vp. These studies of Vp should help to elucidate the physiology of early diastolic LV filling. Further studies are needed to investigate the relationship between factors such as left atrial pressure, severity of mitral regurgitation, and mitral annulus size and Vp.
Declaration of Competing Interests
Drs. Toshihiro Ogawa, Ajit Yoganathan, Cheryl Nordstrom, Julius Gardin, and Renee Bess all declare that they have no competing interests. Dr. David Walker and Mr. Lawrence Scotten state that during the time of the performance of these studies, they were employees of Vivitro Systems, Inc., which was the manufacturer of the in vitro flow model (Superdup'r SD1002) used in this study.
==== Refs
Brun P Tribouilloy C Duval AM Iseriu L Megurira A Pelle G Dubois-Rande JL Left ventricular flow propagation during early filling is related to wall relaxation: a color M-mode Doppler analysis J Am Coll Cardiol 1992 20 420 432 1634681
Stugaard M Risoe C Halfdan I Smiseth OA Intraventricular early diastolic filling during acute myocardial ischemia: assessment by multigated color M-mode Doppler echocardiography Circulation 1993 88 2705 2713 8252682
Takatsuji H Mikami T Urasawa K Teranishi J-I Onozuka H Takagi C Makita Y Matsuo H Kusuoka H Kitabatake A A new approach for evaluation of left ventricular diastolic function: Spatial and temporal analysis of left ventricular filling flow propagation by color M-mode Doppler echocardiography J Am Coll Cardiol 1996 27 365 371 8557907 10.1016/0735-1097(96)81240-X
Garcia MJ Smedira NG Greenberg NL Main M Firstenberg MS Odabashian J Thomas JD Color M-mode Doppler flow propagation velocity is a preload insensitive index of left ventricular relaxation: animal and human validation J Am Coll Cardiol 2000 35 201 208 10636281 10.1016/S0735-1097(99)00503-3
Yellin EL Hori M Yorcan C Sonnenblick EH Gabbay S Frater RW Left ventricular relaxation in the filling and nonfilling intact canine heart Am J Physiol 1986 250 H620 629 3963218
Suga H Goto Y Igarashi Y Yamada O Nozawa T Yasumura Y Ventricular suction under zero pressure for filling Am J Physiol 1986 251 H47 55 3728698
Courtois M Kovacs SJ JrLudbrook PA Physiological early diastolic intraventricular pressure gradient is lost during acute myocardial ischemia Circulation 1990 81 1688 1696 2331773
Greenberg NL Vandervoort PM Thomas JD Noninvasive assessment of diastolic intraventricular pressure gradients using color Doppler M-mode echocardiograph Computers in cardiology 1995 Los Angeles, CA: IEE Computer Society Press 465 468
Smiseth OA Steine K Sandbaek G Stugaard M Gjolberg TO Mechanics of intraventricular filling: study of LV early diastolic pressure gradients and flow velocities Am J Physiol 1998 275 H1062 1069 9724314
Ohte N Naria H Akita S Kurokawa K Hayano J Kimura G Striking effect of left ventricular systolic performance on propagation velocity of left ventricular early diastolic filling flow J Am Soc Echocardiogr 2001 14 1070 1074 11696830
Lemmon JD Yoganathan AP Computational modeling of left heart diastolic function: examination of ventricular dysfunction J Biomech Eng 2000 122 297 303 11036551 10.1115/1.1286559
Vierendeels JA Riemslagh K Dick E Computer simulation of intraventricular flow and pressure gradients during diastole J Biomech Eng 2000 122 667 674 11192390 10.1115/1.1318941
Vierendeels JA Dick E Verdonck PR Hydrodynamics of color M-mode Doppler flow wave propagation velocity V (p): a computer study J Am Soc Echocardiogr 2002 15 219 224 11875384 10.1067/mje.2002.115456
Harrison MR Clifton GD Pennel AT DeMaria AN Effect of heart rate on left ventricular diastolic transmitral flow velocity patterns assessed by Doppler echocardiography in normal subjects Am J Cardiol 1991 67 622 627 2000796 10.1016/0002-9149(91)90902-W
Appleton CP Influence of incremental changes in heart rate on mitral flow velocity: assessment in lightly sedated, conscious dogs J Am Coll Cardiol 1991 17 227 236 1987230
Oniki T Hashimoto Y Shimizu S Kahuta T Yajima M Numano F Effect of increasing heart rate on Doppler indices of left ventricular performance in healthy men Br Heart J 1992 68 425 429 1449930
Yu Ch M Sanderson JE Right and left ventricular diastolic function in patients with and without heart failure: effect of age, sec, heart rate, and respiration on Doppler-derived measurements Am Heart J 1997 134 426 434 9327698
Schober KE Fuentes VL Effects of age, body weight, and heart rate on transmitral and pulmonary venous flow in clinically normal dogs Am J Vet Res 2001 62 1447 1454 11560276
Cheng CP Freeman GL Santamore WP Constantinescu MS Little WC Effect of loading conditions, contractile state, and heart rate on early diastolic left ventricular filling in conscious dogs Circ Res 1990 66 814 823 1689618
Galderisi M Benjamin EJ Evans JC D'Agostino RB Fuller DL Lehman B Levy D Impact of heart rate and PR interval on Doppler indexes of left ventricular diastolic filling in an elderly cohort (the Framingham heart study) Am J Cardiol 1993 72 1183 1187 8237811 10.1016/0002-9149(93)90991-K
Gillam LD Homma S Novick SS Rediker DE Eagle KA. The influence of heart rate on Doppler mitral inflow patterns Circulation 1987 76 IV 123
Steen T Steen S Filling of a model left ventricle studied by colour M-Mode Doppler Cardiovasc Res 1994 28 1821 1827 7867035 7867035
Sugimachi M Sunagawa K Effects of partial left ventriculectomy on left ventricular pump function studied by theoretical analysis J Card Surg 2001 16 24 29 11713853
|
16171516
|
PMC1236943
|
CC BY
|
2021-01-04 16:38:31
|
no
|
Cardiovasc Ultrasound. 2005 Sep 19; 3:30
|
latin-1
|
Cardiovasc Ultrasound
| 2,005 |
10.1186/1476-7120-3-30
|
oa_comm
|
==== Front
Chiropr OsteopatChiropractic & Osteopathy1746-1340BioMed Central London 1746-1340-13-181613892010.1186/1746-1340-13-18ResearchManagement of chest pain: exploring the views and experiences of chiropractors and medical practitioners in a focus group interview Smith Monica [email protected] Dana J [email protected] Robert M [email protected] Palmer Center for Chiropractic Research, Palmer College of Chiropractic, 741 Brady Street, Davenport, IA 52803, USA2005 2 9 2005 13 18 18 29 6 2005 2 9 2005 Copyright © 2005 Smith et al; licensee BioMed Central Ltd.2005Smith 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
We report on a multidisciplinary focus group project related to the appropriate care of chiropractic patients who present with chest pain. The prevalence and clinical management, both diagnosis and treatment, of musculoskeletal chest pain in ambulatory medical settings, was explored as the second dimension of the focus group project reported here.
Methods
This project collected observational data from a multidisciplinary focus group composed of both chiropractic and medical professionals. The goals of the focus group were to explore the attitudes and experiences of medical and chiropractic clinicians regarding their patients with chest pain who receive care from both medical and chiropractic providers, to identify important clinical or research questions that may inform the development of 'best practices' for coordinating or managing care of chest pain patients between medical and chiropractic providers, to identify important clinical or research questions regarding the diagnosis and treatment of chest pain of musculoskeletal origin, to explore various methods that might be used to answer those questions, and to discuss the feasibility of conducting or coordinating a multidisciplinary research effort along this line of inquiry. The convenience-sample of five focus group participants included two chiropractors, two medical cardiologists, and one dual-degreed chiropractor/medical physician. The focus group was audiotaped and transcripts were prepared of the focus group interaction. Content analysis of the focus group transcripts were performed to identify key themes and concepts, using categories of narratives.
Results
Six key themes emerged from the analysis of the focus group interaction, including issues surrounding (1) Diagnosis; (2) Treatment and prognosis; (3) Chest pain as a chronic, multifactorial, or comorbid condition; (4) Inter-professional coordination of care; (5) Best practices and standardization of care; and (6) Training and education.
Conclusion
This study carries implications for chiropractic clinical training relative to enhancing diagnostic competencies in chest pain, as well as the need to ascertain and improve those skills, competencies, and standards for referrals and sharing of clinical information that may improve cross-disciplinary coordination of care for chest pain patients.
Chest PainChiropracticMedical EducationCoordination of Care
==== Body
Background
While the main focus of chiropractic care centers on treatment of musculoskeletal disorders, chiropractors serve as first point of contact with the health care system for patients presenting with a broad range of conditions [1]. As a portal-of-entry healthcare provider in a primary ambulatory setting, the professional responsibilities of the practicing chiropractor include proper assessment, documentation, and treatment of chest pain/discomfort cases, and appropriate and timely referral of chest pain patients as needed.
An extensive body of primary empirical literature addresses patient management protocols (differential diagnosis and diagnostic/treatment algorithms) for patients presenting with chest pain, primarily focusing on cardiopulmonary, gastroesophageal/gastrointestinal, and psychological conditions causing chest symptoms [2-17]. These etiologic sources are ruled out as the cause for many chest pain sufferers, and such patients essentially 'fall out of the algorithm' with ongoing chest pain that remains undiagnosed, untreated, and unresolved.
A small but growing body of literature estimates the presumed prevalence of musculoskeletal chest pain in medical settings at 20–50% [14-18], and reflects a growing awareness that musculoskeletal causes remain largely unexplored as potential sources of chest pain, particularly for chronic or recurrent chest pain that remains undiagnosed and unresolved.
The Cochrane Database of Systematic Reviews (CDSR), containing completed reviews carried out by the Cochrane Collaboration , contains only one citation for chest pain, not musculoskeletal [19]. The Database of Abstracts of Reviews of Effects (DARE), maintained by the NHS Centres for Reviews and Dissemination and linked to the Cochrane Library , includes a number of reviews that focus on comparing various clinical diagnostic test strategies for cardio-related chest pain [20-29], as well as numerous organizational-level studies examining the clinical safety and cost-effectiveness of shifting cardio-related chest pain evaluation units from hospital inpatient to hospital outpatient settings [30-44]. Other review articles returned in the DARE search confirms our impression that current medical approaches to diagnosing, treating, or managing non-specific or non-cardiac chest pain focus principally on psychological and gastroesophageal/gastrointestinal causes and essentially ignore the potential for musculoskeletal etiologies [45-51].
We report on a multidisciplinary focus group project, one aspect of which specifically addressed issues related to the appropriate care of chiropractic patients who present with chest pain, whether as a main presenting complaint or as a co-morbid condition. The prevalence and clinical management, both diagnosis and treatment, of musculoskeletal chest pain in ambulatory medical settings, was explored as the second dimension of the focus group project reported here. The objective was to gain insight into the care and management of patients with musculoskeletal chest pain as experienced by both those with chiropractic training, medical training or combined training.
Methods
Data collection
Focus Group
This project collected observational data from a multidisciplinary focus group composed of both chiropractic and medical professionals. The goals of the focus group were to explore the attitudes and experiences of medical and chiropractic clinicians regarding their patients with chest pain who receive care from both medical and chiropractic providers, to identify important clinical or research questions that may inform the development of 'best practices' for coordinating or managing care of chest pain patients between medical and chiropractic providers, to identify important clinical or research questions regarding the diagnosis and treatment of chest pain of musculoskeletal origin, to explore various methods that might be used to answer those questions, and to discuss the feasibility of conducting or coordinating a multidisciplinary research effort along this line of inquiry.
Population, setting, timeframe
The convenience-sample of five focus group participants included two chiropractors, two medical cardiologists, and one dual-degreed chiropractic/medical physician. The focus group was conducted in early 2004 at the offices of the medical cardiologists.
Support documents/instruments
The questions posed to focus group participants are provided in Additional file 1. Aside from presenting the semi-structured questions, running the audio recorders, and ensuring that all questions were addressed within the time allotted for the focus group meeting, the facilitator's role during the focus groups session was intentionally minimalized in order to enhance the authenticity of the observations offered by focus group participants.
Human subjects
The Institutional Review Board of Palmer College approved this study of human subjects, including the Informed Consent document signed by all Focus Group participants. To protect subject confidentiality, subject records (i.e., signed Informed Consents and verbatim unblinded master transcript) were maintained in a locked file cabinet. The final 'blinded' transcript (all subject identifiers removed) was used during the content analysis, which was performed by all three investigators, although two of the study investigators were also present during the focus group.
Data management and analysis
Focus Group Qualitative Analyses
The focus group was audiotaped and transcripts were prepared of the focus group interaction. Content analysis of the focus group transcripts were performed to identify key themes and concepts, using categories of narratives. All three authors analyzed complete transcripts and developed independent lists of overall themes and concepts subsumed within the general themes. Once completed, the investigators came together to collapse their lists of themes into one set of themes as reached via consensus. This process involved examining themes for commonality, classifying them for uniformity, and then reaching agreement on the final list of six key themes. Once the themes were set and subordinate concepts identified, each investigator looked for quotes and comments which exemplified those themes and concepts (which are presented in the Results, below)
As a methodological 'cross-check', the investigative group's consensus process confirmed observations drawn from each investigator's independent analysis of the transcripts, which strengthened the validity and reliability of the study findings reported from this qualitative research [52,53]. It is important to note that this research is an exploration into the specifics of the convenience sample drawn for the project; therefore, generalizability is not a significant consideration in this study.
Results
Six key themes emerged from the analysis of the focus group interaction, including issues surrounding (1) Diagnosis; (2) Treatment and prognosis; (3) Chest pain as a chronic, multifactorial, or comorbid condition; (4) Inter-professional coordination of care; (5) Best practices and standardization of care; and (6) Training and education. These thematic issues are summarized below, and key excerpts from the focus group transcript exemplifying these thematic issues are included in Additional file 2.
(1) Diagnosis
Participants reported that a good history and physical exam are essential and important to good diagnosis, that a history should include all prior care received for that condition, that records of prior care should be obtained directly from the source provider, and that history, exam, and differential diagnosis are central to the provision of portal-of-entry primary care as well as secondary specialty care. They noted that diagnostic uncertainty, complexity, and discriminant variability are characteristic of chest pain assessment and diagnostic tests, that the inherently high risk of chest pain determines the order of differential workup and the path of diagnostic referral care (e.g., rule out cardiac and other major medical conditions first), and that musculoskeletal chest pain is principally a diagnosis by exclusion. Anecdotal experience of both chiropractic and medical cardiology focus group participants confirms reports in the literature of a high prevalence of suspected musculoskeletal chest pain in ambulatory practice settings.
(2) Treatment and prognosis
Chiropractic participants reported anecdotal evidence (their personal practice experience) of the effectiveness of manual/manipulative approaches to resolve chest pain of suspected musculoskeletal origin. Chiropractic and medical participants both noted lack of formal clinical studies examining effectiveness of manual/manipulative approaches to manage (diagnose and treat) musculoskeletal chest pain, and lack of evidence supporting effectiveness for medical drug interventions for musculoskeletal chest pain (e.g., oral nonsteroidal anti-inflammatory drugs or steroid injections into chest wall), and that it is unknown to what extent drug interventions are prescribed for such conditions in actual current medical practice (generalist or specialist). The agreed that both effectiveness and safety concerns should direct the appropriateness and order of trying various clinical approaches to resolve musculoskeletal chest pain in a given patient, and that a better understanding of the etiology of musculoskeletal chest pain condition(s) would also help discriminate between different conditions and guide the search for identifying effective interventions for a given condition. Natural history or prognosis of treated versus untreated acute or chronic musculoskeletal chest pain is also unknown.
(3) Chest pain as a chronic, multifactorial, or comorbid condition
It is unknown to what extent chronic, unresolved chest pain may represent undiagnosed musculoskeletal chest pain, or to what extent patients with undiagnosed and unresolved musculoskeletal chest pain are perhaps being misclassified as psychological or psychiatric cases. The participants commented that chronic recalcitrant chest pain is associated with high resource use and unsatisfied, distressed patients, that it is unknown to what extent early manual/manipulative intervention in acute musculoskeletal chest pain may prevent development of chronic musculoskeletal chest pain, that chronic musculoskeletal chest pain may raise patient care issues similar to other chronic conditions (i.e., providers and patients may manage some chronic conditions, rather than resolve them), and that the diagnostic and treatment considerations are further complicated when musculoskeletal chest pain and non-musculoskeletal chest pain may exist together as related or unrelated comorbidities. Finally, they noted that with a higher likelihood of degenerative musculoskeletal disorders in older patients and also higher likelihood of visceral (cardiopulmonary or gastrointestinal) disorders in older patients, chest pain in older patients therefore may be more likely of multi-factorial etiology and more likely associated with comorbidities.
(4) Inter-professional coordination of care
Participants reported that referrals should be based on evidence of efficacy/effectiveness for a given condition such as musculoskeletal chest pain, that the path of referral for chest pain will depend on the nature of the condition and the urgency of the situation, that the point of referral may depend on the familiarity or relationship between the providers, and that the nature of the referral (e.g., amount and type of information accompanying the referral) may depend on the nature of the condition, whether the referral is for reasons of diagnosis and/or treatment, the preference of the provider, and the relationship between the providers. Medical specialists (e.g., cardiology) who receive referrals from primary medical practitioners will most typically return the patient to the primary medical practitioner rather than referring them elsewhere, although this also may depend on the nature of the condition and the relationship between the specialty and primary medical practitioner.
Participants felt that patients with co-morbidities (e.g., having both musculoskeletal and non-musculoskeletal chest pain) may be more likely to receive concurrent care from more than one provider, that providers can pro-actively improve interprofessional relationships by being diligent about sharing pertinent information and reports during referrals. Participants felt that educating other providers about available evidence, recognizing and addressing issues of professional boundary protection (often referred to as 'turf'), and that patients' direct experience (with successful or unsuccessful treatment outcome) and their preferences will also impact provider perceptions and interprofessional relationships.
(5) Best practices and standardization of care
Participants reported that standardizing care within professions may facilitate opportunities for interprofessional referrals, that guidelines and care standards are an issue for all professions, that interactions between providers and professions (e.g. referrals) may also be standardized, and that 'best practices' for coordinating musculoskeletal chest pain care would center on the role of primary medical practitioners rather than specialist medical practitioners.
(6) Training/Education
Competencies in exam, diagnostic, and clinical decision-making skills for chest pain were raised as issues for, and by, both chiropractors and medical practitioners. Medical practitioners' perception, familiarity and comfort with chiropractors' diagnostic skills largely comes via direct exposure in postgraduate practice (exchanging clinical reports, etc.) rather than during their medical training. Participants commented that there is a perception that medical education/training is more standardized than chiropractic, and a perception that medical practice is more consistent with medical training (i.e., chiropractors' clinical practice may be more likely to deviate from what they were taught). A comment was made that medical training includes developing skills/competencies in referral practices (e.g., standardized referral forms are used in medical academic practice and teaching clinics).
Discussion
The focus group dialogue suggested several implications for current and future chiropractic practice, undergraduate and post-professional chiropractic education and clinical training, research, and professional organization or policy. These implications for practice, education, research, and policy are summarized below along with our recommendations.
Clinical practice
With all portal-of-entry providers such as chiropractors, the responsibility to diagnose chest pain is vital. The focus group touched on this point several times. In order to arrive at a diagnosis for chest pain, or any other condition, they stressed the importance of first taking a good history and then performing a thorough examination. The diagnosis of chest pain, however, is complicated and requires excellent diagnostic skills. The focus group (both the chiropractors and the medical practitioners) expressed a concern over the ability of chiropractors to accurately diagnose chest pain. In order for chiropractors to have a role in managing chest pain from the point of entry, they must acquire and demonstrate competence in diagnosing the complaint.
Chest pain can have a multitude of etiologies, involves an inherently high level of diagnostic uncertainty, and diagnostic algorithms are complex. The clinician's most immediate concern is ruling out emergent versus non-emergent conditions [54-58]. The focus group participants were unified in voicing the need for rapid diagnosis and management for cardiopulmonary conditions such as myocardial infarction and pulmonary embolism, among others. They also pointed out that the clinical picture of chest pain can be complicated by simultaneous etiologies. For example, one of the medical doctors noted that in his own practice he saw patients with cardiac disease and chest wall tenderness. This sentiment was echoed by one of the chiropractors who noted that simply palpating a patient's chest wall and finding tenderness does not rule out cardiac or other life threatening causes of chest pain. Therefore, a full chest pain work up must include evaluations for cardiac, pulmonary, gastrointestinal, musculoskeletal and psychological causes of chest pain.
Once life threatening causes of chest pain have either been ruled out or managed, other possible etiologies may be investigated. The focus group expressed concern that musculoskeletal chest pain may be either missed or misdiagnosed as psychological in nature. The misdiagnosis of musculoskeletal chest pain as psychological could cause much distress, cost, and unnecessary suffering for patients. It is important, therefore, to investigate efficient and accurate diagnostic strategies for this complaint.
Participants in the focus group commented that musculoskeletal chest pain is common in their practices, both chiropractic and medical. This is consistent with reports in the literature that 20%–50% of chest pain presentations in ambulatory settings may be musculoskeletal [14-18]. Management of chest pain of cardiac or gastrointestinal origin is much more standardized than musculoskeletal chest pain. Appropriate protocols and treatment algorithms do not exist for musculoskeletal chest pain. Manipulation, physiological therapeutics, injections, analgesics, and other treatments may be employed, but none have been extensively investigated.
The opportunity for cross-disciplinary coordination of care for chest pain certainly exists. Unfortunately, effective referral pathways do not exist. Chest pain in medical practice is often diagnosed by cardiologists who then send patients back to their referring clinicians, most likely a primary care medical physician. Therefore, it is important to ascertain and possibly improve those skills, competencies, and standards for referrals and sharing of clinical information that may improve current and future cross-disciplinary coordination of care for chest pain patients.
Clinical and health services research
It is apparent that there is insufficient scientific evidence to guide clinical practice and decision making for musculoskeletal chest pain. The focus group chiropractic participants largely reported only personal anecdotal evidence for the effectiveness of manipulative interventions for chest pain of suspected musculoskeletal origin, and both the chiropractors and medical practitioners commented upon this lack of evidence, both in terms of therapy as well as for diagnosis. Similarly, there is a concomitant lack of evidence supporting the chemotherapeutic interventions used by medical doctors for suspected musculoskeletal chest pain.
Thus, there is much that is not known. Research questions worth investigating include:
• What is the incidence and prevalence of musculoskeletal chest pain in chiropractic clinical practice?
• What is the incidence and prevalence of musculoskeletal chest pain in specialist cardiologist practice? In general medical practice?
• What percentage of chiropractors treat musculoskeletal chest pain compared to those who refer out for care?
• How effective is manipulation for treating musculoskeletal chest pain?
• What other modalities do chiropractors use during such treatments?
• What diagnostic methods are used for determining the presence of musculoskeletal chest pain? What is the reliability, validity, sensitivity and specificity of each test?
• What are the costs involved in treating such cases?
• What interdisciplinary models exist with regard to developing coordinated-care protocols for diagnosis and treatment of acute musculoskeletal chest pain? For long-term management of chronic or recurrent musculoskeletal chest pain?
• Do incidence and prevalence rates vary geographically or by setting?
One challenge relative to chiropractic research is that funding sources are limited and few opportunities exist. So, this presents a conundrum, in that more research is needed but the greatest amount of resources (both funding, and limited research workforce) are directed toward conditions which are already well established with regard to chiropractic research: low back pain, neck pain, and headache.
A Search using the key terms "Chiropractic" and "Musculoskeletal Chest Pain" on PubMed yielded only three papers, two of which had no real pertinence to the issue at hand. The third paper was a case report that looked at using a specific chiropractic adjusting procedure for managing chronic chest pain [59]. There were no randomized trials found in the literature. Shifting the search to the terms "Medicine" and "Musculoskeletal Chest Pain" improved the yield to just nine papers, one of which was a repeat from the chiropractic search, and several of which were tangential to this issue. Obviously, this is an area needing far more research.
As a first pass at documenting and better understanding this problem area, it would be useful to survey the chiropractic profession to quantify rates of musculoskeletal and non-musculoskeletal chest pain presentations in clinical practice, whether as a chief complaint, or as a related or unrelated comorbid condition. The incidence rate of chest pain presentations to chiropractic teaching clinics has been estimated to range from 1% to 7% [60,61], but rates in a typical chiropractic practice are essentially unknown. In surveying chiropractors regionally or nationally, it would also be worthwhile to compare incidence and prevalence rates in rural versus non-rural chiropractic practice, given that chiropractors located in rural or underserved areas may be more likely to serve as the patient's first contact with the health care system, or to function as their patient's main or usual source of care in a broader generalist capacity compared to chiropractors in other areas [62,63]. Chiropractors serving as a first contact or portal-of-entry in a primary care setting may be more likely to see chest pain cases presenting earlier during an episode of care-seeking, or more likely as a generalist to serve as a main source of care overall for an entire chest pain episode. Along that same line, a comparison of rates in chiropractic versus medical practice may also provide insight as to potential implications for management and co-management of these conditions and patients in the primary care setting.
It is important to note that such research should of necessity be collaborative and interdisciplinary. As noted, funding opportunities within chiropractic are limited, yet chiropractors are working in collaboration with medical physicians in a variety of settings. Following case reports and case series which suggest a role for manipulative intervention in musculoskeletal chest pain, the next step would be to devise collaborative research within medical settings, acknowledging that this is likely the best location to obtain participants for research. A multi-disciplinary practice-based research effort could provide a foundation for conducting the requisite feasibility studies and generating the necessary preliminary clinical data and methods (e.g., developing protocols and establishing reliability of procedures), that can then guide and justify more extensive, more rigorous, controlled preclinical and clinical trial work along this line of inquiry.
Education and educational research
What becomes obvious is that a lack of research has impact and implications for the education of both chiropractors and medical physicians with regard to managing chest pain of musculoskeletal origin. This is also the case with inter-professional collaboration and referral. One comment made by a participant (the dual qualified chiropractor/medical practitioner) was that most medical physicians do not know enough about the training of chiropractors, and do not know about chiropractic diagnostic acumen.
In responding to the question asked by one participant as to how relations between chiropractors and medical practitioners might change, one of the chiropractic physicians answered "better education" and noted that the Council on Chiropractic Education has laid out what he referred to as "the minimum requirements" for education in this field. This suggests that not enough is being done to enhance the education of chiropractic students with regard to musculoskeletal chest pain, and perhaps it is necessary to undertake a study across the chiropractic colleges.
Chiropractic colleges in North America tend to use chiropractic physicians as the main faculty in the diagnosis classes. Therefore, it is commonly the case that cardiology classes are taught by chiropractic physicians with expertise in family practice, such as those who have earned diplomate certification (advanced postgraduate training programs) in family practice and internal medicine. However, these programs should not be seen as equivalent to medical residencies in family practice. They usually require approximately 360 additional hours of didactic training, with only part of that training in a clinical setting. There are efforts underway to develop the 'advanced practice' chiropractic physician, with one chiropractic educational institution offering a training program coordinated by a medical physician. However, this is in its infancy and much needs to be worked out. This does indicate a growing level of interest in this kind of training, and will produce chiropractic physicians even better able to correctly differentiate a diagnosis of musculoskeletal chest pain. The participants in the focus group made no mention of postgraduate training or opportunity, yet this represents the one area beyond the standard curriculum where further training can be gained and is therefore the only way in which current chiropractors may finally gain new understanding of processes such as discussed here.
It would be worthwhile to survey the academic institutions to gain a better understanding of what is currently being taught across the chiropractic curriculum, as well as through postgraduate offerings. Focus groups comprised of diagnosis and cardiology instructors, postgraduate instructors, and representative field practitioners, can be performed. Such efforts would help derive a better picture of the current reality across the chiropractic profession with regard to education in musculoskeletal chest pain.
Interestingly, concern was also raised by participants that the current training in medical schools was inadequate, and that in part this was due to instructors who were not completely conversant with some of the more traditional means of diagnosis. An excerpt from the medical focus group comment underscores this observation: "... most cardiology now is very simple... Which test will give me the diagnosis? Most cardiologists...don't know auscultation... how to listen to the heart... professors for 15–20.... were never taught." This comment suggests that the 'low-tech' art of auscultation is being lost in medical practice, or increasingly replaced by 'higher-tech' laboratory testing in driving the process of diagnosis for cardiac problems, an opinion echoed in the literature [64,65]. A study of curricular and postgraduate medical education may similarly provide insights relative to current teaching and skills development for medical practitioners in managing musculoskeletal chest pain.
Profession and health policy
As a professional issue, attention needs to be afforded the inherent uncertainty and complexity of chest pain diagnosis and the sometimes dynamic interplay between 'diagnosis' and 'treatment' whether in managing a given patient in actual practice or in attempting to define an appropriate evidence-based professional 'standard of care'. This is perhaps particularly true for a condition such as musculoskeletal chest pain, given the obvious dearth of established evidence from which the clinician may draw or on which to form definitive professional recommendations to guide current clinical practice. As a corollary example of an empirical 'treatment-based' diagnostic strategy, a presumptive diagnosis of gastroesophageal reflux disease (GERD) may be pragmatically validated in practice following a patient's favorable symptomatic response to a short course of prescriptive therapy such as proton-pump inhibitors, perhaps preempting or potentially avoiding more invasive or costly diagnostic tests such as endoscopy [66-72]. Non-cardiac chest pain, defined most simply as recurrent episodes of unexplained retrosternal pain in patients lacking a cardiac abnormality after a reasonable evaluation, is associated with repeated emergency room utilization [73] and may be treated empirically with antidepressants [74].
We might ask whether there is a role for musculoskeletal assessments within clinical chest pain diagnostic algorithms that is not being fully exploited in current practice, particularly given the relative costs and safety of the more invasive and resource-intensive alternatives, not to mention patient preferences. In cases where an early 'low-tech' assessment offers a presumptive suggestion that chest pain may be musculoskeletal in nature, might a short course of manual therapy help to validate a presumptive diagnosis and guide treatment decisions [75-80]? How much, and what level, of current available evidence is needed to support clinical decisions or professional recommendations that favor low-technology, low-cost noninvasive procedures early in the diagnostic workup, or that justify manual therapy following an empirical validation of a presumptive musculoskeletal diagnosis? These are tough questions with no easy answers, especially given the inherent high-risk, uncertainty, and complexity of chest pain diagnosis, and the possibility that chest pain may present with any mix of multiple musculoskeletal or non-musculoskeletal etiologies or comorbidities.
Musculoskeletal chest pain as an area of inquiry fits well within the health services research agendas outlined in health policy initiatives to improve primary care, patient safety, and the delivery of evidence-based cost-effective care. As identified earlier, the appropriate management of chest pain raises a host of considerations relative to improving cross-disciplinary coordination of care within the health care system, whether for diagnostic consult, referral for treatment, or continuity of care in the overall management of the patient's total care plan. The potential implications for improving patient safety are also worth noting, specifically relative to enhancing prompt accurate diagnosis, and where possible decreasing unnecessary exposure of patients to higher risk or more invasive procedures. As a specific target within the primary care, patient safety, and cost-of-care initiatives, 'ambulatory-care sensitive conditions' are identified as those conditions that, when managed appropriately in the outpatient setting, can prevent unnecessary and costly inpatient care. Chest pain is high on the list of high prevalence ambulatory conditions associated with 'avoidable hospitalizations', and with repeated high-cost hospital emergency room utilization [73]. While discipline-specific approaches to diagnosing or treating non-cardiac chest pain of gastrointestinal, psychiatric, or musculoskeletal origin have served useful, the quality and safety of patient care may be even better served by a coordinated cross-disciplinary research effort and practice approach.
A final health policy consideration relative to health workforce planning and development, is in acknowledging that chiropractors serve a role as a first point of contact with the health care system or as the main source of care for many patients, particularly in rural or medically underserved areas [62,63]. Policies to improve access to care by promoting the primacy of the relationship between usual-source practitioners and their patients, must also pay due attention to developing the role and requisite skills of non-medical practitioners such as chiropractors to appropriately manage or co-manage a broad range of conditions such as chest pain in primary care settings.
Conclusion
Our research leads us to offer a number of recommendations for practice, research, education, and policy. Certainly, the investigators and members of the focus group feel that more education should be required in the diagnosis and management of chest pain. Research is also needed about the educational opportunities and challenges revolving around interdisciplinary care and practice.
Greater outreach to the medical research community, and indeed to the wider medical community, will help to enhance skill sets and collaborative opportunities. This outreach may help to drive research in those areas where it is most needed: diagnosis, incidence/prevalence, treatment, and clinical protocols within and across disciplines. By developing a research base, it will be possible to establish appropriate standards for care, and these can be enhanced by creating multidisciplinary panels to explicitly improve cross-disciplinary coordination of care.
Competing interests
The author(s) declare they have no competing interests.
Authors' contributions
Monica Smith conceived the study, and coordinated the focus group meetings as well as performed thematic analysis of transcripts and helped write the manuscript. Dana Lawrence performed thematic analysis and coding of transcripts and prepared components of the manuscript. Robert Rowell also performed thematic analysis and coding of transcripts and prepared components of the manuscript. All three authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Focus Group Questions for MD & DC Chest Pain Study.
Click here for file
Additional File 2
Seminal excerpts of dialogue from focus group transcripts, bytopic.
Click here for file
==== Refs
Christensen MG Kerkhoff D Lollasch MW Cohn L Job Analysis of Chiropractic 2000 Greeley: National Board of Chiropractic Examiners
Gustafson DH Risberg L Gering D Case studies from the quality improvement support system Agency for Health Care Policy and Research (AHCPR) 1997, publication no 97-0022 9 16
Wax CM Abend DS Pearson RH Chest pain and the role of somatic dysfunction JAOA 1997 97 347 355 9232945
Roberts RR Zalenski RJ Mensah EK Cost of an emergency department based accelerated protocol vs. hospitalization in patients with chest pain JAMA 1997 278 1670 1676 9388086 10.1001/jama.278.20.1670
Weingarten SR Riedinger MS Conner L Practice guidelines and reminders to reduce duration of hospital stay for patients with chest pain: an intervention trial Ann Int Med 1994 120 257 263 8291818
Hamm CW Goldmann BU Heeschen C Kreymann G Berger J Meinertz T Emergency room triage of patients with acute chest pain by means of rapid testing for cardiac troponin T or troponin I N Engl J Med 1997 337 1648 1653 9385123 10.1056/NEJM199712043372302
Tatum J Jesse R Kontos MC Comprehensive strategy for the evaluation and triage of the chest pain patient Ann Emer Med 1997 29 116 125
Mikhail M Smith F Gray M Britton C Frederiksen S Cost-effectiveness of mandatory stress testing in chest pain center patients Ann Emer Med 1997 29 88 98
Zalenski R Rydman R McCarren M Feasibility of a rapid diagnostic protocol for and emergency department chest pain unit Ann Emer Med 1997 29 99 108
Graff LG Dallara J Ross MA Joseph AJ Itzcovitz J Andelman RP Impact on the care of the emergency department chest pain patient form the chest pain evaluation registry (CHEPER) study Am J Cardiol 1997 80 563 568 9294982 10.1016/S0002-9149(97)00422-0
Lee TH Pearson SD Johnson PA Failure of information as an intervention to modify clinical management: a time-series trial in patients with acute chest pain Ann Int Med 1995 122 434 437 7856992
Lee TH Juarez G Cook EF Ruling out acute myocardial infarction: a prospective multicenter validation of a 12-hour strategy for patients at low risk N Engl J Med 1991 324 1239 1246 2014037
Pozen MW Agostino RB Selker HP Sytkowski PA Hood WB A predictive instrument to improve coronary-care-unit admission practices in acute ischemic heart disease N Engl J Med 1984 310 1273 1278 6371525
Brush JE Brand DA Acampora D Chalmer B Wackers FJ Use of the initial electrocardiogram to predict in-hospital complications of acute myocardial infarction N Engl J Med 1985 312 1137 1141 3920520
Svavarsdóttir AE Jonasson MR Gudmundsson GH Fjeldsted K Chest pain in family practice: diagnosis and long-term outcome in a community setting Can Fam Phys 1996 42 1122 1128
Selbst SM Ruddy RM Clark BJ Henretig FM Santulli T Pediatric chest pain: a prospective study Pediatrics 1988 82 319 323 3405660
Selbst SM Chest pain in children Pediatrics 1985 75 1068 1069 4000782
Klinkman MS Stevens D Gorenflo DW Episodes of care for chest pain: a preliminary report from MIRNET J Fam Pract 1994 38 345 352 8163958
Swingler GH Zwarenstein M Chest radiograph in acute lower respiratory infections in children The Cochrane Database of Systematic Reviews 2003 CD001268
Heidenreich PA Go A Melsop KA Prediction of risk for patients with unstable angina Agency for Healthcare Research and Quality 2000 31 Rockville MD Evidence Report/Technology Assessment
Garber AM Solomon NA Cost-effectiveness of alternative test strategies for the diagnosis of coronary artery disease Ann Intern Med 1999 130 719 728 10357690
Kuntz KM Fleischmann KE Hunink MG Douglas PS Cost-effectiveness of diagnostic strategies for patients with chest pain Ann Intern Med 1999 130 709 718 10357689
Raggi P Callister TQ Cooil B Russo DJ Lippolis NJ Patterson RE Evaluation of chest pain in patients with low to intermediate pretest probability of coronary artery disease by electron beam computed tomography Am J Cardiol 2000 85 283 288 11078293 10.1016/S0002-9149(99)00733-X
Heller GV Stowers SA Hendel RC Clinical value of acute rest technetium-99 m tetrofosmin tomographic myocardial perfusion imaging in patients with acute chest pain and nondiagnostic electrocardiograms J Am Coll Cardiol 1998 31 1011 1017 9562001 10.1016/S0735-1097(98)00057-6
Kosnik JW Zalenski RJ Grzybowski M Huang R Sweeny PJ Welch RD Impact of technetium-99 m sestamibi imaging on the emergency department management and costs in the evaluation of low-risk chest pain Acad Emerg Med 2001 8 315 323 11282665
Muttreja MR Mohler ER Clinical use of ischemic markers and echocardiography in the emergency department Echocardiography 1999 16 187 192 11175140
Shaw LJ Heller GV Travin MI Cost analysis of diagnostic testing for coronary artery disease in women with stable chest pain J Nucl Cardiol 1999 6 Economics of Noninvasive Diagnosis (END) Study Group 559 569 10608582 10.1016/S1071-3581(99)90091-0
Ben Gal T Zafrir N The utility and potential cost-effectiveness of stress myocardial perfusion thallium SPECT imaging in hospitalized patients with chest pain and normal or non-diagnostic electrocardiogram Isr Med Assoc J 2001 3 725 730 11692545
Underwood SR Godman B Salyani S Ogle JR Ell PJ Economics of myocardial perfusion imaging in Europe – the EMPIRE Study Eur Heart J 1999 20 157 166 10099913 10.1053/euhj.1998.1196
Tosteson ANA Goldman L Udvarhelyi IS Lee TH Cost-effectiveness of a coronary care unit versus an intermediate care unit for emergency department patients with chest pain Am Heart Assn 1996 94 143 150
Gomez MA Anderson JL Karagounis LA Muhlestein JB Mooers FB An emergency department-based protocol for rapidly ruling out myocardial ischemia reduces hospital time and expense: results of a randomized study (ROMIO) J Am Coll Cardiol 1996 28 25 33 8752791 10.1016/0735-1097(96)00093-9
Robinson DJ Woods PG Snedeker CA Lynch JH Chambers K A comparison trial for stratifying intermediate-risk chest pain: benefits of emergency department observation centers Prev Cardiol 2002 5 23 30 11872988
McManus RJ Mant J Davies MK A systematic review of the evidence for rapid access chest pain clinics Int J Clin Pract 2002 56 29 33 11833553
Dougan JP Mathew TP Riddell JW Suspected angina pectoris: a rapid-access chest pain clinic Qual J Med 2001 94 679 686
McCullough PA Ayad O O'Neill WW Goldstein JA Costs and outcomes of patients admitted with chest pain and essentially normal electrocardiograms Clin Cardiol 1998 21 22 26 9474462
Mikhail MG Smith FA Gray M Britton C Frederiksen S Cost-effectiveness of mandatory stress testing in chest pain center patients Ann Emergency Med 1997 29 88 98
Goodacre S Morris F Arnold J Angelini K Is a chest pain observation unit likely to be cost saving in a British hospital? Emerg Med J 2001 18 11 14 11310454 10.1136/emj.18.1.11
Goodacre SW Morris FM Campbell S Arnold J Angelini K A prospective, observational study of a chest pain observation unit in a British hospital Emerg Med J 2002 19 117 121 11904256 10.1136/emj.19.2.117
Zalenski RJ Grzybowski M The chest pain center in the emergency department Emerg Med Clin North Am 2001 19 469 481 11373990 10.1016/S0733-8627(05)70195-4
Roberts RR Zalenski RJ Mensah EK Costs of an emergency department-based accelerated diagnostic protocol vs hospitalization in patients with chest pain JAMA 1997 278 1670 1676 9388086 10.1001/jama.278.20.1670
Bing ML Abel RL Sabharwal K McCauley C Zaldivar K Implementing a clinical pathway for the treatment of Medicare patients with cardiac chest pain Best Pract Benchmarking Healthcare 1997 2 118 122
Weingarten S Ermann B Bolus R Early "step-down" transfer of low-risk patients with chest pain: a controlled interventional trial Ann Internal Med 1990 113 283 289 2115754
Caragher TE Fernandez BB Barr LA Long-term experience with an accelerated protocol for diagnosis of chest pain Arch Pathol Lab Med 2000 124 1434 1439 11035571
Ng SM Krishnaswamy P Morissey R Clopton P Fitzgerald R Maisel AS Ninety-minute accelerated critical pathway for chest pain evaluation Am J Cardiol 2001 88 611 617 11564382 10.1016/S0002-9149(01)01801-X
Kisely S Campbell LA Skerritt P Psychological interventions for symptomatic management of non-specific chest pain in patients with normal coronary anatomy The Cochrane Database of Systematic Reviews 2005 CD004101 15674930
Ofman JJ Dorn GH Fennerty MB Fass R The clinical and economic impact of competing management strategies for gastro-oesophageal reflux disease Aliment Pharmacol Ther 2002 16 261 273 11860409 10.1046/j.1365-2036.2002.01167.x
Borzecki AM Pedrosa MC Prashker MJ Should noncardiac chest pain be treated empirically? A cost-effectiveness analysis Arch Intern Med 2000 160 844 852 10737285 10.1001/archinte.160.6.844
Fass R Empirical trials in treatment of gastroesophageal reflux disease Dig Dis 2000 18 20 26 10729734 10.1159/000016930
Botoman VA Noncardiac chest pain J Clin Gastroenterol 2002 34 6 14 11743240 10.1097/00004836-200201000-00004
Fass R Fennerty MB Ofman JJ The clinical and economic value of a short course of omeprazole in patients with noncardiac chest pain Gastroenterology 1998 115 42 49 9649457
Eslick GD Coulshed DS Talley NJ Review article: the burden of illness of non-cardiac chest pain Aliment Pharmacol Ther 2002 16 1217 1223 12144570 10.1046/j.1365-2036.2002.01296.x
Lincoln YS Guba EG Denzin N, Lincoln Y Paradigmatic controversies, contradictions, and emerging confluences Handbook of Qualitative Research 2000 2 Thousand Oaks CA Sage Publications
Lincoln YS Guba EG Naturalistic inquiry 1985 Beverly Hills CA; Sage Publications
Ornato JP Evaluating the patient with chest pain Patient Care 2001 35 54 71
Mootz RD Talmage DM Clinical assessment strategies for the thoracic area Top Clin Chiropr 1999 6 1 19
Souza TA Differential diagnosis and management for the chiropractor 2002 2 Gaithersburg: Aspen Publishers
Haneline MT Chest pain in chiropractic practice J Neuromusculoskeletal System 2000 8 84 8
Wells KA Averting disaster – a case report of overlooked angina pectoris in a chiropractic setting J Neuromusculoskeletal System 2000 8 89 97
Polkinghorn BS Colloca CJ Chiropractic management of chronic chest pain using mechanical-force, manually-assisted short-lever adjusting procedures J Manipulative Physiol Ther 2003 26 108 115 12584509 10.1067/mmt.2003.8
Smith M Ellerbrock M Khorshid K Handley S Retrospective study of chest pain cases presenting to a chiropractic teaching clinic: a preliminary feasibility study J Neuromusculoskeletal System 2000 8 67 75
Smith M DeBono V Retrospective records review to study chest pain in a chiropractic teaching clinic setting: further exploration of feasibility J Neuromusculoskeletal System 2000 8 76 83
Smith AJPH Smith M Carber L Chiropractic health care in Health Professional Shortage Areas (HPSAs) of the U.S Am J Public Health 2002 92 2001 2009 12453823
Hawk C Nyiendo J Lawrence D Killinger L The role of chiropractors in the delivery of interdisciplinary health care in rural settings J Manipulative Physiol Ther 1996 19 82 91 9064315
Shima MA Evaluation of chest pain: back to the basics of history taking and physical examination Postgrad Med 1992 91 155 64 1603748
Mangione S Nieman LZ Cardiac auscultatory skillsof internal medicine and family practice trainees. A comparison of diagnostic proficiency JAMA 1997 278 717 722 9286830 10.1001/jama.278.9.717
Numans ME Lau J de Wit NJ Bonis PA Short-term treatment with proton-pump inhibitors as a test for gastroesophageal reflux disease Ann Internal Med 2004 140 518 527 15068979
Borzecki AM Pedrosa MC Prashker MA Should noncardiac chest pain be treated empirically? A cost-effectiveness analysis Arch Intern Med 2000 160 844 852 10737285 10.1001/archinte.160.6.844
Botoman VA Noncardiac chest pain J Clin Gastroenterol 2002 34 6 14 11743240 10.1097/00004836-200201000-00004
Fass R Empirical trials in treatment of gastroesophageal reflux disease Dig Dis 2000 18 20 26 10729734 10.1159/000016930
Fass R The clinical and economic value of a short course of omeprazole in patients with noncardiac chest pain Gastroenterol 1998 115 42 49
Ofman JJ The clinical and economic impact of competing management strategies for gastro-oesophageal reflux disease Aliment Pharmacol Ther 2002 16 261 273 11860409 10.1046/j.1365-2036.2002.01167.x
Ofman JJ The cost-effectiveness of the omeprazole test in patients with noncardiac chest pain Am J Med 1999 107 219 227 10492314 10.1016/S0002-9343(99)00219-3
Aikens JE Michael E Levin T Myers TC Lowry E McCracken LM Cardiac exposure history as a determinant of symptoms and emergency department utilization in noncardiac chest pain patients J Behavior Med 1999 22 605 617 10.1023/A:1018745813664
Varia I Logue E O'Connor C Randomized trial of sertraline in patients with unexplained chest pain of noncardiac origin Am Heart J 2000 140 367 372 10966532 10.1067/mhj.2000.108514
Christensen HW Vach W Manniche C Haghfelt T Hoilund-Carlsen PF Diagnosis and treatment of musculoskeletal chest pain in patients with suspected stable angina pectoris European J Chiropr 2002 49 92 93
Christensen HW Vach W Gichangi A Manniche C Haghfelt T Hoilund-Carlsen PF Cervico-thoracic angina identified by case history and palpation findings in patients with stable angina pectoris J Manipulative Physiol Ther
Brodsky AE Cervical angina: a correlative study with emphasis on the use of coronary arteriography Spine 1985 10 699 709 4081876
Yeung MC Cervical disc herniation presenting with chest wall pain Can J Neurol Sci 1993 20 59 61 8467431
Wells P Cervical angina Am Fam Physician 1997 55 2262 1164 9149653
Jacobs B Cervical angina NY State J Med 1990 90 8 11
|
16138920
|
PMC1236944
|
CC BY
|
2021-01-04 16:38:23
|
no
|
Chiropr Osteopat. 2005 Sep 2; 13:18
|
utf-8
|
Chiropr Osteopat
| 2,005 |
10.1186/1746-1340-13-18
|
oa_comm
|
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-131613524610.1186/1745-0179-1-13ReviewMigration and mental health in Europe (the state of the mental health in Europe working group: appendix 1) Carta Mauro Giovanni [email protected] Mariola [email protected] Maria Carolina [email protected] Josep Maria [email protected] "Report on the Mental Health in Europe" working group [email protected] Department of Public Health University of Cagliari, Italy2 Unitat de Recerca i Desenvolupament, Hospital Sant Joan de Deu-SSM, Barcelona, Spain2005 31 8 2005 1 13 13 20 3 2005 31 8 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.Background
This paper is a part of the work of the group that carried out the report "The state of the mental health in Europe" (European Commission, DG Health and Consumer Protection, 2004) and deals with the mental health issues related to the migration in Europe.
Methods
The paper tries to describe the social, demographical and political context of the emigration in Europe and tries to indicate the needs and (mental) health problems of immigrants. A review of the literature concerning mental health risk in immigrant is also carried out. The work also faces the problem of the health policy toward immigrants and the access to health care services in Europe.
Results
Migration during the 1990s has been high and characterised by new migrations. Some countries in Europe, that have been traditionally exporters of migrants have shifted to become importers. Migration has been a key force in the demographic changes of the European population. The policy of closed borders do not stop migration, but rather seems to set up a new underclass of so-called "illegals" who are suppressed and highly exploited. In 2000 there were also 392.200 asylum applications.
The reviewed literature among mental health risk in some immigrant groups in Europe concerns: 1) highest rate of schizophrenia; suicide; alcohol and drug abuse; access of psychiatric facilities; risk of anxiety and depression; mental health of EU immigrants once they returned to their country; early EU immigrants in today disadvantaged countries; refugees and mental health
Due to the different condition of migration concerning variables as: motivation to migrations (e.g. settler, refugees, gastarbeiters); distance for the host culture; ability to develop mediating structures; legal residential status it is impossible to consider "migrants" as a homogeneous group concerning the risk for mental illness. In this sense, psychosocial studies should be undertaken to identify those factors which may under given conditions, imply an increased risk of psychiatric disorders and influence seeking for psychiatric care.
Comments and Remarks
Despite in the migrants some vulnerable groups were identified with respect to health problems, in many European countries there are migrants who fall outside the existing health and social services, something which is particularly true for asylum seekers and undocumented immigrants. In order to address these deficiencies, it is necessary to provide with an adequate financing and a continuity of the grants for research into the multicultural health demand. Finally, there is to highlight the importance of adopting an integrated approach to mental health care that moves away from psychiatric care only.
==== Body
1. Migration in Europe: the social, demographic and political context
Trends in migration in Europe
The number of migrants in the world has more than double since 1975, with most living in Europe (56 million), Asia (50 million) and Northern America (41 million) [1]. In 1990, migrants accounted for over 15% of the population in 52 countries. Most of the migration was from developing to developed countries, even though migration within developing countries is also increasing as the pace of economic growth and the demand of labour between these countries also changes [2].
In Europe, mass migration is not new. During the 20 th century Europe has experienced three major periods of movements: around the time of the First and Second World Wars and during last decade. Migration during the 1990s has been high. It is a period characterised by new migrations, especially from the Eastern and Central European countries and from the Commonwealth of Independent States, but it is the Balkans war, which has dominated these movements [3]. Some countries in Europe, like Ireland, Spain, Italy or Portugal, that have been traditionally exporters of migrants have shifted to become importers. Given the growing economic and political cooperation emerging between many European countries, competition for highly-skilled labour will intensify and the demand of more labour movement will surely continue to grow as organisms such as the International Organisation for Migration predicts.
The role of migration in European population change is being debated during recent years as a result of growing concerns about issues such as demographic ageing, shortages of working age populations and payment of pensions among others [4]. The United Nations Population Division has suggested that Europe might need replacement migration to cope with this potential problems, ranging from around one to 13 million new immigrants per year from 2000 to 2050 [5].
Migration has been a key force in the demographic changes of the European population during last decade. Specifically, the components of population change for the period 1997–99, indicate that migration was the most important component in 33 per cent (15 out of 46) of the countries. 33 countries experienced an increase in population during that period, while 13 had a decrease. Of the 33 countries which had an increase in population, 25 had a net gain of population through migration [3]. The total recorded stock of foreign population living in Europe in 1999/2000 was 21.16 million people, constituting a 2.6% of the aggregate population in Europe [3]. This figure might surely have increased during recent years as immigration rates in many countries have not ceased to grow.
Nevertheless, within Europe there exist great differences in the pattern of migration depending on the countries. This way, some experts have come to distinguish a Northern and a Southern model of immigration in Europe. The Northern model would be composed of those Northern European countries (e.g. United Kingdom, Netherlands, Germany and Sweden among others), which have had a long experience of immigration through history, and especially right after Second World War, when arrival of immigrants experienced a sharp increase. The Southern model would comprise Southern European countries for which the immigration phenomenon is relatively recent, such as Spain or Portugal. While in Spain the percentage of foreign 'legal' residents sum up to 3.2% of total population in 2002, in other European countries this percentage fluctuates between 5–10% at this time [5].
The composition of immigration by countries is also very different. The vast majority of immigrants in Central and Eastern Europe and Scandinavian countries come from elsewhere in Europe. Germany's immigration field is also strongly European, and along with Austria and Finland receives a high proportion of its immigrants from Central and Eastern Europe. In contrast, the Mediterranean countries, together with UK and Netherlands, attract a high proportion of immigrants beyond Europe. Almost a third of UK's and Spain's immigrants come from outside Europe [3].
Political and socio-economic instability in and around Europe has significantly increased the number of refugees and asylum seekers arriving in European countries.
A refugee is "one who flees to a foreign country or power to escape danger or persecution" [6]. Although the UNHCR statistics are based on this definition, many agencies reserve the term 'refugee' for those whose application for asylum under the terms of the Refugee Convention has been accepted. This is to distinguish them from 'asylum seekers', who still have to prove their right to asylum, and 'illegal aliens', who may be fleeing from danger or persecution, but have not entered the official asylum procedure or have been rejected by it.
According to United Nation High Commission for Refugees (UNHCR) in 2000 there were 392.200 asylum applications in Europe. Northern European countries have been receiving for a longer period of time a great population of refugees, while countries from Southern Europe receive a vast majority of economic migration. However, the low numbers of asylum seekers in Southern European countries may be misleading owing to the large amount of refugees that are thought to by-pass the step of applying for asylum, then getting the residence permit through immigration laws or becoming "illegal aliens" [7].
The presence of undocumented immigrants is a well-established fact in most European countries. They come or are "called" into Europe to perform badly paid, physically and psychologically stressful jobs in highly qualified service economies and welfare states. The closure of frontiers to new immigration has not prevented the increase of "illegal aliens" in Europe. On the contrary, the welfare gap between Europe and its neighbouring countries makes such jobs more attractive for enterprising women and men in the poorer and poorest regions of the world. According to the International Labour Office estimates, in 1991 there were around 2.6 million of undocumented immigrants living in Europe [3]. This is the last official data estimated. Other sources nowadays estimate the number of undocumented immigrants in Europe in more than 3 million [8], finding the largest populations in the Southern European countries due to their structure, culture, history and geography [5]. One of the features of this structure, present in the South of Europe, is the great expansion of the black economy. Although there is no way to confirm these numbers, the trend seems to go upward.
Policy and political context of migration in Europe
Trends in migration in Europe commenced to transform a few decades ago as the result of changes that were taken place in the economic, political and social realms. This inflection point was described by Massey et al [9] to take place during the 1970s, when patterns of migration meant a rupture to those of all past migration movements. Among the new features of these new migrations we find, on behalf of the immigrants, the fact that they are not simply looking for an immediate job but also for situations which give them a higher quality of life, together with better future possibilities for themselves and their children; on behalf of the reception countries, the development of a labour legislation before unknown which have given place to a clear contradiction between legal and illegal work [10]. These trends, which have changed the profile and perception of international migration that marked the post war era, have become especially visible during the 1990s. Due to this new model of migration, characterised for the change into a post-industrial society, there is an urgent need for the elaboration of policies that respond to this new global situation.
EU member states have been practicing a policy of closing borders throughout the 1990s, a policy that have been specially hardened during recent years. Different strategies for restricting the entrance of immigrants have been developed by the different countries. For example, in Belgium, Italy and Spain, the states have mixed restrictive immigration (e.g. establishing a quota system for work-related migration) policies with recurrent processes of regularisation of undocumented immigrants, while in Germany, the state has opted for restricting immigration via the mechanism of political asylum, ignoring in this way the increasing number of undocumented immigrants http://www.freudenbergstiftung.de.
However, the policy of closed borders do not stop migration, but rather seems to set up a new underclass of so-called "illegals" who are – against all declarations of human rights – inhumanely suppressed and highly exploited. Inconsistency and the lack of a future vision have marked the immigration and asylum policies of most European countries for the last fifteen years.
The EU is still in its way to elaborate a common policy and therefore, determine its immigration model. For that the 7th Conference of Minister responsible for Migration Affairs was held in Helsinki during September 2002. Nevertheless, some sources point to the possible election of a model of (work-related) quotas, a model that has been widely criticized because of its proved failure up to now. An active and common policy of immigration not only focused on control but also on integration policies, is needed in Europe, since immigration has never stopped and will continue.
2. Needs and (mental) health problems of immigrants
Among all the changes a human being must face throughout his live, few are so wide and complex as those which take place during migration. Practically everything that surrounds the person who emigrates changes. Aspects ranging from diet, family and social relations to climate, language, culture, and status are subject to change. The decision to migrate originates in a perceived lack of prospects that a person has in his own country. Every person who emigrates experiences affective loss, but is buoyed up in the hope of finding the first world paradise they often know so little about.
The singularity of the migratory experience lies in the fact that it is a psycho-social process of loss and change, which is known in the psychiatry of migration as a grief process [11]. The grief is considered as a type of stress characterised by its intensity and length [12]. The process of migration has been explained through a model comprised of seven griefs (losses) causing anguish that a person will experience with time: family and friends, language, culture, homeland, loss of status, loss of contact with the ethnic group, and exposure to physical risks. Furthermore, the migratory grief process is partial, as the grief's subject do not really disappear (unlike the grief for the death of a loved one), and recurrent, as the contact with the country of origin could revive the bonds [13]. Difficulties in expressing grief can cause psychological problems. These difficulties are accentuated when migration is accomplished under adverse conditions. The reception in the new country is crucial for the complete and successful elaboration of the grief process.
In the case of refugees, who have to flee their country for fear of being persecuted, the grief process is more complex. There is the fact that they can not go back to their country of origin, so that their grief is, in this case, closer to the experience of loss, than to that of separation. War-related experiences and occupational status before migration have been proved to be risks factors for psychiatric symptoms among Somali refugees living in the UK [14]. As for every immigrant, the post-migration environment has been proved to have a considerable influence regardless of prior traumatic exposures, being the level of affective social support in exile an important determinant of the severity of possible disorders [15].
As a diagnosed condition Post Traumatic Stress Disorder (PTSD) is by far the most common mental health problem among refugees and asylum seekers [16], followed by mood disorders. According to the Medical Foundation in the UK, the range of psychological problems experienced by torture survivors can include: nightmares, hallucinations, panic attacks, sexual problems, phobias, difficulty in trusting others and forming relationships and depressive illness or anxiety [16].
A study carried out in Portugal with asylum seekers and refugees found that principal sources of stress of this population were connected to traumatic or harrowing experiences in the country of origin due to political or family reasons, problems of communication (language), legal status and work.
In Spain, researches [17-19] have been carried out on mental and psychosomatic disorders in immigrant populations. These researches found the following factors associated with the immigrants' mental health: labour and economic instability, cultural and social marginalisation, family estrangement, pressures to send money to their families, racial discrimination and lack of statutory documentation [20].
In spite of these considerations, which seem to evidence that it is impossible to consider "migrants" as a homogeneous group concerning the risk for mental illness according to Murphy's hypothesis [21], many studies seem to confirm that not a generalized psychopathological risk is present among emigrants. In this sense, psychosocial studies should be undertaken to identify those factors which may, under given conditions, imply an increased risk of psychiatric disorders and influence seeking for psychiatric care [22]. Specific migrant situations may be evaluated, concerning psychopathological risk and seeking for psychiatric care, in order to variables as: motivation to migrations (see Rack work subdividing settler, refugees, gastarbeiters, [23]; distance for the host culture (religion, language, and so); ability to develop mediating structures [24]; legal residential status.
The Chronic and Multiple Stress Syndrome (Ulysses syndrome)
The particularly hard conditions of today's migration seem to be propitiating a worsen in the mental health of the newcomers. Current situations are making of the migratory experience, an extremely hard and unbearable process. Thousands of people are emigrating and suffering levels of stress, in some cases, inhuman. Examples of this, it's the situation given in the South of Europe particularly Spain and Italy, where in recent years, the coasts of Sardinia, Sicily, Southern Spain and the Canary Archipelago have registered a massive arrival of small, fragile crafts which hundreds of Africans use to get into European territory putting their lives in danger [20].
Psychiatrists from the Psycho-pathological and Psycho-social Assistance Service (SAPPIR) team, located in Barcelona, have described the common symptoms that most immigrants present when attending the centre and have called it Chronic and Multiple Stress Syndrome in immigrants (or Ulysses syndrome), relating the risky and hard journey that the immigrants pursue in search of a better life with the odyssey of the mythical Greek character in his long voyage through the Mediterranean. Immigrants affected by this syndrome present depressive symptomatology with atypical characteristics, where depressive symptoms are mixed with anxious, somatoform and dissociative symptoms [13]. The development of this condition occurs progressively as the immigrant encounters the obstacles that take place during the migration process: dangerous journey, distance from their own environment and family, difficulties to find a job, food, housing and obtain documents, and the racism suffered in the reception country. The psychiatrist's team propose that the 'Chronic and Multiple Stress Syndrome' should constitute an autonomous category situated in between adjustment disorders and Posttraumatic stress disorder [13].
The growing incidence of this syndrome in many psychiatrist and psychological services across Europe have alerted a group of social scientists and health care professionals from different countries, who have addressed to the EU Parliament, in order to make an urgent call about the situation which it has been worsening during last five years [8].
3. Epidemiological studies
The first epidemiological evaluations of psychiatric disorders in immigrants date back to more than seventy years ago [25]. However, the results obtained in these studies were somewhat limited due to a number of methodological problems which have only recently been overcome thanks to the introduction of standardized diagnostic criteria and structured interviews and to investigations being carried out on the general population and not only on subjects who had been seen by health professionals [26].
Even so, currently in Europe epidemiological studies which offer information on mental health status of immigrants are still very scant. Most studies carried out until nowadays are clinical and mostly limited to private medical practices, as epidemiological research offers by now little information. These clinical studies have some limitations due to the use of small and, frequently, biased samples.
There is little data at our disposal with regard to the level of psychic and physical health among those who are culturally different owing to inadequate systems of registration (e.g. lack of the 'nationality' variable). These systems have recently started to be modified in some countries in order to provide soon with useful data on the health profile of the newcomers.
Nevertheless, some epidemiological studies do exist although its distribution across UE countries is very patchy. For example, while in some countries such as the United Kingdom, Netherlands and Sweden there are some data concerning the mental health status of immigrants (including in this group asylum seekers and refugees), in Belgium there are no nationwide or regional epidemiological studies available on immigrants, as neither, for example, in Spain or Portugal.
The following review – clustered by countries – underline the different trend of research at present time in evolution in each European country. Some topics as drug consumption in immigrated people and psychiatric health in refugees were summarized in a transnational way. Some of these studies were highlighted in the Council of Europe's report on Health Conditions of migrants and refugees in Europe [27]. The more relevant research concerning Refugees from Balkans was from USA (Harvard programme from refugees), it was summarized considering the relevance of this health problem for the European Union.
- United Kingdom
• Hospital admission rates for schizophrenia are highest for people with of Caribbean, Irish and Polish background [28-32] with special high incidence in 16–29 year old people of Caribbean background. Also higher rates than the indigenous are found in people from Indian and Pakistan. Two studies collecting rates of first onset schizophrenia were recently carried out in Trinidad and London [33], the results confirm that sending countries have low rates of schizophrenia. Author indicates that the impact of migration itself produces high stress but rates of schizophrenia are even higher in the second generation, suggesting that other social factors may be responsible for the increase. Social factors and genetic vulnerability may play a role considering that only sub groups of migrants show higher rates of psychosis. A review of the literature about seventeen population based studies [34] proposed that the developmental task for formulating the life plan challenges the young adult's executive function abilities, which may be weaker in individuals vulnerable to schizophrenia. Formulating the life plan may be made more difficult by the position of disadvantaged ethnic minorities, raising the risk for schizophrenia. A more recent review [35] indicates that African-Caribbean population in England is at increased risk of both schizophrenia and mania and the excess of the two psychotic disorders are probably linked: African-Caribbean patients with schizophrenia show more affective symptoms, and more relapsing course with greater social disruption but fewer chronic negative symptoms, than White patients.
• Suicide rates of young women immigrants from the Indian subcontinent are consistently higher than those of their male counterparts and of young women in the indigenous populations of the countries to which they immigrated [36,37]. Family conflict appears to be a precipitating factor in many suicides, whereas mental illness is rarely cited as a cause. Depression, Anxiety and domestic violence may contribute to the high rates. Authors suggest that affective disorders may be underdiagnosed in this population.
• The issue concerning the lower rate of recognized mental disorders in women of Indian origin was addressed by a survey of Jacob and coll. carried out in a general practice setting in West London [38]. Common mental disorders were documented in 30% of patients (similar in Indian women to those in other UK populations), individuals with common mental disorders had a higher frequency of consultation, were less likely to see depression as an indicator for medical intervention and were more likely to withhold some of their concerns from the general practitioner (GP). Incorrect diagnosis by the GP was most likely to occur when patients did not disclose all their complaints. Authors conclude that differing conceptualisation of common mental disorders may contribute to their under-recognition in women of Indian origin. A recent survey examined the contribution of Asian ethnicity in rates of practice prescribing for antidepressant and anxiolytics medication in 164 general practices in East London [39]. Results indicate that where the proportion of Asian patients is high, both antidepressants and anxiolytics prescribing is low. The study confirms that the putative low rates of non-psychotic disorders attribute to Asian population is not a selective feature of access to secondary care but uncertainty remains as the low rate of antidepressants and anxiolytics prescription of GP may be due to a lower prevalence rate, or to "culture-bound syndromes" or to practical difficulties in diagnosis and management within the general practice setting.
• The alcohol abuse among people of Indian descent is reflected in rates of cirrhosis-related mortality, which are twice as high as among English males [40]. Drug abuse has also been reported as a problem [41].
- The Netherlands
• According to UK studies a series of researches conducted in the Netherlands indicate that the incidence [42] and point prevalence [43] of schizophrenia is increased in several (Morocco, Suriname and Netherlands Antilles) but not all (Turkish and Western countries) immigrants groups. The relative risk of schizophrenia in Suriname born immigrants against Suriname resident born population was 1.46 [42]. Authors conclusion is that selective migration according to Odegaard's selection hypothesis cannot solely explain the higher incidence of schizophrenia in Surinamese immigrants to the Netherlands.
• Drug abuse related schizophrenia together with psychotic diseases was found not uncommon among some migrant groups [44], especially migrants from Suriname and the Antilles. It has been suggested as an explanation for the increased incidence of psychosis among some immigrant groups. A recent population-based, first contact incidence study found misused of illicit substance in: 23% of Dutch, 17% of Moroccans, 27% of Surinames and 30% of Turkish. These results seem to suggest that a higher rate of drug misuse is an unlikely explanation for the increased incidence of psychotic disorders among Moroccan and Surinamese immigrants in the Netherlands [45].
• In a country where the unemployment rate among migrants was almost three times that of the nationals in 1994, the suicide rate among children of immigrants was considerably higher than that of the national population [46].
• A recent study carried out in randomly selected sample of Turkish immigrants, found a prevalence rate of "minor" psychiatric disorders (33.4%, 36.1% in females, 27.9% in males), higher than those normally found in community based samples. The results suggest that the expression of somatic complaints around "tightness" should alert physicians to further explore symptoms of minor psychiatric disorders and to examine sources of distress stemming from the partner relationship, the family, the work and from the poor housing and financial conditions [47].
• A study on service utilisation in women immigrants in Amsterdam found that Surinamese, Antillean, Turkish and Moroccan women made considerably less use of mental health care services than native born women. Immigrant women consulted, on the contrary, social work facilities and women crisis intervention centres nearly 1.5 times more than mental health care services [48]. Exploring the reason for the ethnic difference in services utilisation authors conclude that cultural and socioeconomic factors are largely responsible of such a difference. The results imply that a care policy may improve the accessibility of mental health services for immigrant women. They also suggest to employ more ethnic and bilingual care providers. Cultural barriers in detecting problem behaviour in children was explored by a study of Crijnen and colleagues [49]. Turkish immigrant teachers reported high levels of anxiety and depression in immigrant Turkish children which go largely undetected by their Dutch teachers.
- Italy
• A community survey using the standardised clinical interview Present State Examination [24,50] demonstrated that Senegaleses travelling salesmen living in Sardinia, whose working conditions facilitate a community lifestyle, do not appear to be at risk for depression when compared to Sardinian controls. Unexpectedly, higher rates of anxiety and depressive disorders were shown in the few fellow-countrymen who had managed to obtain a steady job with regular wages. In the latter case, the onset of psychopathological disorders was closely associated with the loss of contact with fellow-countrymen. A sample of Moroccan emigrants employed in similar occupations was characterised by a higher risk respect to a population of Senegalese salesmen, and an increased incidence of psychopathological episodes in the six months following their arrival in Sardinia. The authors argued how elements of cultural cohesion such as those represented by the associations of Islamic confraternities may exert strong protective factors against the development of psychopathological, particularly depressive, disorders in immigrants from Senegal.
• A study evaluated the frequency of ICD-10 psychiatric disorders in a community sample of subjects of Sardinian origin resident in Paris. Results were compared to data obtained from a sample of the general Parisian population and from a sample of Sardinians resident in Sardinia [51]. Migration was shown to be associated with a higher risk both of anxiety (as people living in Sardinia) and depressive disorders in the young people (as Parisians). The young emigrants and the children of emigrants (2nd generation emigrants) seem to be prone to drug-abuse and bulimia. The presence of a confidential relationship appears to have a protective effect as shown by the much higher incidence of depression in emigrants who are widowed, separated or live alone. This suggests the need for support strategies.
• Elderly Sardinian residents who had experienced migration are characterised by an increased risk of dysthymia [52]. The problems of adjustment of the return migrant, particularly in the elderly is presented in some Italian papers [53,54]. The focus is on immigration from southern Europe and Turkey toward northern European countries and on progressive aging of people migrated in the 50s and 60s. Authors affirmed that little is known about the health of migrants once they return to their country of origin but the issue represents a very relevant health problem. A recent community survey found a higher frequency of depressive disorders in the Sardinian immigrants in Argentina [55]. Female (not male) showed a higher risk in respect to Sardinians resident in Sardinia. Results seem to suggest that previous emigration in a country that decreased dramatically their economical level may predispose subjects for depressivedisorders comparing with their native population and with subjects migrated in countries more economically advantaged because the lifetime rate of depressive episodes in the Sardinian immigrants in Paris, as reported in a previous above cited research, was lower than in Sardinian immigrants in Argentina. The study suggests the need of systematic researchers and support for European citizen migrated to south America and other economically disadvantaged countries.
- Germany
• Hansen and coll. [56] attempt to evaluate if the elevated rate of schizophrenia among migrants has been explained in part by possible misdiagnosis measuring the extend of misdiagnosis among Turkish immigrant patients and German controls. Three investigators (a researcher of Turkish origin, a German researcher and a clinician) evaluated independently two sample of patients one of Turkish an one of German origin. The rate of potential misdiagnosis was higher among migrant, yet not strongly correlated to poor secondary language proficiency. The same research group [57] found in a Turkish of schizophrenic patients a higher rate of depression and hostile excitement than in German schizophrenic patients. Authors say that such a figure may be mainly due to diagnostic differences.
• To solve the problems of language and culture barriers that raise problems for the diagnostic and therapeutic process, Grube developed a bilingual setting by integrating a Turkish psychologist who belongs to a counselling centre into a therapeutic team [58]. Comparing outcome measures in Turkish patient against non Turkish migrants as controls, results showed shorter periods of hospital treatment for Turkish non-schizophrenic in-patients and Turkish schizophrenic patients achieve higher level of rehabilitation. Turkish patients with all kinds of psychiatric diagnoses showed lower ratios of readmission.
• A study on psychiatric inpatient in Frankfurt a.M. found suicidal attempts more frequent among the Mediterranean girls than among their German counterparts [59].
- Sweden and Scandinavian countries
• The issue about increasing risk for schizophrenia and severe mental disorders in immigrants was studied by several Nordic authors. Hjern and coll [60] found in one of the first longitudinal studies analysing incidence rates of first hospital admission, that five and second generation immigrants have an increased risk for severe psychiatric disorders compared to natives. The highest risks were found in Finnish first-generation immigrants. First generation immigrants from refugee countries have lower risk for alcohol or substance abuse than natives. However the risk in second second-generation refugees is significant higher than in natives. First generation immigrants from refugee countries have the same risk for schizophrenia as native while the risk in second generation refugees is higher than in natives. A complex model integrating both the "Goalstriving stress hypothesis" (the level of psychological stress which derived from a discordance between the goals to be reached and the degree of aspiration) [61] and the cited hypothesis of Eaton and Harrison (concerning executive function abilities and developmental task for formulating the life plan challenges which may be weaker in individuals vulnerable to schizophrenia and which may be made more difficult by the position of disadvantaged ethnic minorities, raising the risk of schizophrenia) [34] may be taken in account. The role of ethnic factors in the risk of schizophrenia was underlined in a study carried out in Malmo [62]. This work confirmed that compared with those who were native-born, migrants had increased risk for Schizophrenia like psychosis. But the risk was most markedly increased in immigrants from East Africa. On the contrary background factors as extreme duress of migration did not appear contribute strongly to increasing risk.
• Also data from Danish Civil Registration System seem to confirm the association between migration and schizophrenia both in immigrants to Denmark and among Danish with a history of foreign residence.
• A paper explored structures of illness meaning among somatizing-Turkish-born migrant women living an a poor and low status suburb of Stockholm [63]. Migrants communicated distress by concrete expression about the body, emotion, social and life situation. Pain was prominent and often lateralised to one side of the body. The use of traditional expression of distress ranged from open use of avoidance. Attribution wascharacterised by verbalising links of coherence between health and aspect of life. Psychiatric attribution was rarely accepted or valued as tool for recovery, or as helpful in linking bodily symptoms to emotional distress. The results of this study point out the mutual need of exploring meaning in the clinical encountered to help patients, particularly migrant, make sense out of different perspectives of illness and healing.
• Compared with Swedish immigrant from Iran, Chile, Turkey, Poland Kurdistan (but for this group differences being not significant) had an increased risk of self reported longstanding psychiatric illness and for intake of psychotropic drugs. Living alone, poor knowledge of the Swedish language, non employment, and low sense of coherence were strong risk factor for self reported longstanding psychiatric illness and for intake psychotropic drugs [64]. Country of birth was a significant risk for poor self reported health and psychosomatic complaints in women of reproductive age in Sweden. Swedish born (but not Finnish) women and female refugees reported more psychosomatic complaints in the 90s than in 80s [65]. Similarly to the cited study on Sardinian immigrants to Paris, the results do not appear to confirm the clinical findings of "somatization" as a privileged "psychopathological course" in latin immigrants reported in the past [66].
• The alcohol related disorders in immigrants was studied by a register based work on a national cohort of 1.25 million youth born 68–79 and 1.5 million of adults born 1929–65 [67]. First and second generation immigrants from Finland (but not from Southern and middle East Europe and non European countries) had a higher relative risk for hospital admission because of an alcohol related disorder compared to the Swedish population. Authors say that patterns of alcohol abuse in the country of origin are strong determinants of alcohol related disorders in first generation immigrants. The patterns in second generation immigrants are influenced by parental countries of origin as well as pattern in the majority of population. The Finnish minority and intercountry adoptees are of particular concern in prevention. Intecountry adoptees have also a high risk for severe mental health problems, suicide and suicide attempts [68].
• A registry study in the county of Copenhagen indicated an excess of Mental Retardation in children from ethnic minorities. The cause are not known and not are aetiological factors for Mental Retardation for a great part of the children. Consanguinity is likely to be a risk factor for Mental Retardation [69].
- Spain
• Excluding the data from obstetrics and paediatrics, depressive disorders were the second cause of medical consultations in "undocumented" immigrants in the district of Usera-Villaverde, Madrid [70]. In a random sample of immigrant psychiatric patients, depression was more frequent than in the control group of native psychiatric patients, and alcohol abuse less frequent [71]. Among the clinical studies carried out in Spain, there is the annual report of the SAPPIR (Psycho-patological and Psychosocial assistance service for immigrants). The centre focuses its research by symptoms instead of diagnosis. This is due to the fact that in most cultures of origin of immigrants who visit the service, emotion is expressed through the body, but without alexithymia. This approach, which integrates mind and body, favours that depressive symptomatology come along with/is accompanied by somatic symptomatology. According to SAPPIR's research findings, the most frequent symptomatology in immigrants is the triad insomnia – fatigue – migraine [72].
- Belgium
• Findings of a study carried out in Brussels [73] conformed increased incidence of psychosis in second generation Moroccan immigrants to Belgium.
- France
• French papers on migration studied more frequently descriptive psychopathology about interpretative psychology than epidemiology. Several authors who had studied the emigration of Islamic populations in France, indicated how many individuals were affected by a certain degree of conflict respect to their migrant situation, with particular concern for the dispositions laid down by the Koran which prohibit leaving the country of origin for non-religious reasons [74]. Problems and conflicts concerning the failure of integration of Moslem French immigrants, particularly Algerians who chose French nationality in 1962, has been taken up in 50 case histories reported by Pouget and coll. [75].
• More recently two cross-sectional studies on perceived health were performed in the Comoro Island and in South Eastern of Marseille. Immigrants in Marseille have higher perceived health status than those living in Comoro Islands for the dimension of physical, mental and general health. The perceived health score level of migrated people is closed to those reported in France [76].
- Ireland
• Ireland became only recently a destination for migration, thus studies about immigration are still rare. One of the last report about emigration for Ireland, studying attitudes toward emigration found men who contemplated emigration reporting higher self-esteem scores, and women contemplating emigration reporting lower self esteem scores. Women who contemplated emigration had higher depression scores than women who did not contemplate emigration. This pattern was not evident for men. The results indicate that psychologically women view the prospect of emigration less positively than men [77].
• These different attitudes toward emigration in men and women suggest a heuristic hypothesis explaining the different risk in migration (gender determined) shown in the above mentioned Sardinian studies. That hypothesis can be taken into account if similar attitudes will be demonstrated in another catholic but Mediterranean culture. The study on Sardinian immigrants in Argentina suggests that emigration to a country which subsequently suffers dramatic economic problems may increase risk for depressive disorders (but only in women) which are not found in emigrants to economically stable countries. In the survey on Sardinian immigrants in Paris, only the lifetime rate of Depressive Episode in young men immigrants in Paris was higher than the lifetime rate of Depressive Episode in young men Sardinian resident in Sardinia.
• It is very important to underline the two different diametrically opposite economic conditions in which the two migratory waves occurred between Paris and Argentina. Paris offered successful work opportunities but also a new world based on competition. Even though France is closer to Sardinia than Argentina there was a risk, specially for young people, to lose their traditions and family ties [78].
• The hypothesis that arises from the data analysis of the Sardinian and Irish study is that migrated women, with a possible more frequent "depressive" coherent interpretation of self and knowledge [79] may be higher involved during difficult situations when the family safety is under the threat of economical instability. Men may be more at risk in situations of rapid improvement where the competitive challenge becomes pressing as the risk of "goal striving stress" [61].
• The hypothesis of men sensitivity to goal striving stress may be applied to understand the way Ireland with best European economic performance during the period 1980–2000 (the only European country with both increased mean income and decreased unemployment rate) present a suicide male increase around 130%. During the last twenty years period suicides decreased in all Europe but not in the Irish men, instead, the Irish female suicide rate decreased.
• Consistent with constructivist cognitive concepts [79,80], a "depressive style of knowledge" would imply a system whereby the individual explicitly attempts to maintain a coherent image through application of a theory which contemplates a pessimistic view of the world and the future, assuming a sort of "defeat-oriented hyper-responsibility", tacitly challenging the losses experienced.
• Paradoxically depressive attitudes, in face to stress conditions of social change may exert a protective factor if the "compulsive self-responsibilisation" towards the newly-originated socially successful opportunities may modulate the search of new roles which are able to provide subjects with an adequate income and with a sense of leadership whilst at the same time to maintain a role which is more socially accepted and which is perceived as being more "traditional". Several interesting lines of psychosocial research have in fact hypothesised that the cultural transmission tends to be perpetuated when the particular cultural institutions (in this case the social role) are perceived by the individual (and by the other members of the group) as an integral part of the evolving self, but which at the same time, are able to meet new needs and requirements [81]. The latter hypothesis was elicitated to explain the absence of conflicts and the lower risk for depression in a sample of women employed as nurses (a innovative role but that preserve traditional roles) in an African society in rapid change [82].
- Portugal
• Literature on immigration in Portugal is at present inconsistent. Most international papers are still interested in immigration from Portugal. The study of Ferron and Coll. in Switzerland [83] found in labour migrant adolescents (from Portugal, Spain, Italy, Turkey and former Yugoslavia) lower health behaviour (notably sexual behaviour and substance abuse) than Swiss counterpart. Only alcohol consumption and drink driving place migrant adolescent in a lower risk than their Swiss peer.
• An example of an interdisciplinary approach to provide a framework for understanding and promote mental health in the context of cultural diversity in a Canadian perspective for Portuguese immigrant is presented in a paper from James and Prilleltensky [84]. The framework is not only limited to assessing the needs of individual but draws on anthropology, philosophy, political science, and religious studies to understand the social, cultural, moral, and religious domains. The need for observation in the specific group ethnic model predicting depression in immigrant women (with specific attention to Portuguese) in an a flexible, individualized approach to ethnic women's psychology health care is suggested by Franks and Faux in their Ontario Study [85].
- Greece
• More than ten years ago a series of very relevant papers appears in the international literature about immigration from Greece. Literature on immigration in Greek is at present inconsistent. The work from Mavreas VR and Bebbington PE [86] shows that the rates of psychiatric disorders in two Greek sample (one of which of Greek Cipriots living in Camberwell, London and the other one living in Athens) were somewhat higher than those of the Camberwell population. Greeks reported more symptoms of general anxiety disorders. This is consistent with the results others researches and with the Sardinian immigration studies in which a higher frequency of anxiety disorders in the Sardinian sample and of depressive disorders among Parisians was observed. Mavreas and Bebbington suggest a greater risk of anxiety disorders in southern and of depression in northern European countries.
• A paper from Fitcher and coll. [87] found significantly higher GHQ-28 scores for Greeks and Turks adolescents in their homeland as compared to Greek adolescent in Munich. Results seems do not confirm the acculturation stress hypothesis.
- Austria
• A paper describes the behaviour disturbance and emotional problems of Turkish immigrant and Austria children aged 9 to 13 years living in Vienna, rated by their parents and both by Turkish and German teachers. The prevalence of behaviour problems did not differ between the Turkish and Austrian children [88].
• A paper presented at the DTGPP Satellite Conference, Ethnicity & Mental Health in Europe, Essen Germany, September 2003 [89] deals with the issue of the misdiagnosis of schizophrenia in migrants, authors think that hebephrenic schizophrenia in migrants may besometimes misdiagnosed as chronic depressive status due to cultural barriers.
- Drug abuse
• Reports to investigate the reasons for drug abuse among immigrant youth have been carried out in Sweden [90], France [91-94] and Germany [95] coming up with similar conclusions which suggest that drug abuse was a consequence of difficult social integration. A 1996 WHO report noted that the consumption of tranquillisers and antidepressants by young immigrants across Europe is growing. Alcohol abuse is also reported as an important problem among immigrants population in Europe.
• A recent review of the literature underlined that the association between migration and addiction is very heterogeneous. More or less drug and alcohol dependence than native population have been reported in different migration phenomena across the world [96] As suggested in some above cited studies on alcohol abuse, but probably not with the same strong association, patterns of addiction abuse in the country of origin are determinants of alcohol related disorders in first generation immigrants. The author of the cited lecture indicates, in spite of the chief role of migration drug problems in public debate and concern, a lack data about alcohol dependence in the migration population in Europe. The report of the European Union Strategy on Drugs 2000–2004 has no mentions or suggestions of these specific problems in any European Union country.
- Refugees and mental health
• Recent surveys have shown that two thirds of refugees experience anxiety and/or depression [97].
• Refugees have a high incidence of post traumatic stress disorder, depression, anxiety, panic disorder and agoraphobia [98].
• Shortages of food, being lost in war situation, being close to death and suffering serious injury were each related to specific psychiatric symptoms in a community sample of adult Somali refugees [14].
• A relevant work on psychiatric disorders on Bosnian refugees was performed by the Harvard USA program in Refugee Trauma [99,100] The initial study reported a high rate of disabling depression and post traumatic stress disorder among refugees. Bosnian Refugees followed up initially found asylum in Croatia in 1996 and now returning in Bosnia or remained in the Balkans (70.4%9) or migrated in European Union or USA (21.3%), the others had died (21.3%) or were lost to follow-up [100]. Nearly 50% former Bosnian refugees who remained living in the Balkan area still present psychiatric symptoms and disability 3 years after initial assessment. Depression was unremitting, disabling, and potentially associated with premature death in the elderly. About 20% of those did not have symptoms of psychiatric disorder at starting time had symptoms at follow-up. Depressed refugees had three times the risk of dying than non depressed. Refugees who emigrated were more traumatised, better educated and had fewer health and mental health problems than the majority who remained in the region.
• A recent lecture [46] on global risk and protective factors in the development of mental disorders in refugees cutting across lines of social class and cultural identity, analysing a community based study on Iraqui refugees in The Netherlands, suggests that a long asylum procedure is associated with psychiatric disorders. And indicates that both policy makers and mental health workers should take note of this findings.
ESEMeD
The recent Study of the Epidemiology of Mental Disorders – ESEMeD, a survey carried out in six European countries (France, Belgium, Italy, Spain, The Netherlands, Germany), could mean an important contribution for the study of epidemiology of mental disorders in immigrants living in Europe. The ESEMeD's sample consisted in a random selection of 21.425 individuals from the general population aged 18 or older and non-institutionalised. This sample comprises a 6.03% (6.58%weighted) of foreign individuals. The study will provide us with highly relevant information concerning prevalence and comorbidity of mental disorders, and patterns of service utilization of these populations.
The recent Report "The State of Mental Health in Europe" [101] compared the ESEMED findings for psychological distress, as measured by the SF-12 questionnaire, for people who were born in the country compared to those who not born in the country. Because of the way the sample were designed, it was not possible to compare those born in the country with those not born in the country for Italy. The country ratios (risk of distress in immigrants against natives) were: 1.0 in Belgium, 1.2 in The Netherlands, 1.4 in Spain, 4.3 in Germany, 4.7 in France.
4. Health policy toward immigrants: access to health care services
Despite in the migrants groups vulnerable populations were identified with respect to health problems, in many European countries there are migrants who fall outside the existing health and social services, something which is particularly true for asylum seekers and undocumented immigrants (due to the fact that most asylum applications are refused, the division line between both situations is very small). They are usually only entitled to emergency health services.
For example, exclusionary policies in relationship of access to health services on behalf of asylum seekers have been documented in France, where limitations in access were found for those asylum seekers who are not granted a work permit, even if they can't work due to ill health or language barriers [102].
In spite of some states efforts to universalise the right to access national health care services, therefore including undocumented migrants there exist an important lack between rights and accessibility [20]. Undocumented immigrants may not access health care services for causes such as administrative obstacles or fear of being reported to the police. This action is not in practice nowadays in most countries, but nevertheless certain evidence of it exists. According to the study Easy Scapegoats: Sans Papier immigrants in Europe [103], the German Law of Foreigners obligates public institutions to denounce undocumented people to the Office of Foreigner's Affairs. Some hospitals have informed the police while treating undocumented immigrants.
Other obstacles immigrants in general may find when attempting to seek for health care in European countries despite their legal situation are the lack of adequate information about the available health care facilities and communication (language) problems. Moreover, in the case of psychological problems, it may result an obstacle the fact that mental health problems may lead to stigmatisation of that group.
Nevertheless, the health care gaps that are being left by the authorities are being covered by the informal work of doctors at the health system and by NGO's, which provide with medical and, specially, mental health assistance together with health promotion and prevention programs among other services.
5. State of the art in Mental Health Care provision
Researches in the field of access to care deal with subjective and objective barriers. The objective barriers either lie in the domain of the available information, the structure of the health care system, as well as the availability of treatment modalities. The subjective barriers lie either on the side of affected person, the treatment professionals orhealth care planners. Ethnic minority groups, particularly migrants, are faced with several potential barriers in the access of care, resulting in a lower representation in mental health services [56].
Taking the above into account, an examination of studies of mental health services for migrants groups in Europe shows that quality and availability of this provision is very patchy between countries. For example, a mapping exercise of mental health care in Europe carried out by Watters [104] suggested a serious absence of mental health care for migrants in Central and Eastern Europe, something of great importance if we take into account the rapid economic and social changes affecting these regions [104].
Delivering of specific mental health care services for migrants is integrated into mainstream mental health services, who provide the majority of these, in some countries (Netherlands, United Kingdom) even though there exist parallelly specific community services mainly funded by the government. In other countries, these specific community services and NGOs are the main providers of mental health services for these groups while the national health system adapts to the new health needs in a very slow manner (Spain, Portugal). Nevertheless, even those countries with a longer experience of migration and where the bulk of mental health services are being provided by the statutory sector, denounce the weakness of this provision. Most important weaknesses are reported next.
Communication problems are especially significant as a factor that generates exclusion and have been reported in all countries studied. Language is the most essential work instrument of the mental health care provider as the analytical and therapeutic processes are heavily dependent on language as a medium. Drawing on recent studies, services are almost always only available in the majority language [104,7]. The Health for all, all in health study [104] reported that most migrants surveyed tended to think that care providers underestimate the language problems and also say that language difficulties make them more aggressive and paranoid towards the care provider.
The diagnostic process is complicated by the cultural differences between the immigrant and the therapist. Diagnostic mistakes are more often than with native patients. It is important to consider the immigrant's ethnic identity as well as his lay models explaining his illness. Furthermore, his expectations about the treatment must be considered if the treatment shall be successful. Integration of immigrants into a developed network of psychiatric care system may be a useful strategy. Another condition for a more effective psychiatric care of immigrants is a well foundedtranscultural psychiatric training of the professionals [105].
There is a clear deficiency in training and education practices on the subject of understanding immigrant culture for mental health professionals working with migrants, as every country makes the demand of this special training. If well it is true that some countries of Northern Europe have a major and longer experience in this realm, Fernando [106] stresses the necessity of urgent change in the training of professionals, the way mental health assessments are made and the narrow eurocentric nature of therapy. It has been reported that, in many cases, psychosocial problems can mislead the health care staff who are not familiar with the process and impact of migration on psychological health [107].
On the other hand, there exists a big demand for the assistance of interpreters or health agents – who are under utilised – so that migrants are not force to rely on friends and family members to act as interpreters, a practice which has been described by the head of one transcultural psychiatric unit as unethical and totally unacceptable [104]. However, immigrants often feel discomforted because they suspect the interpreters to disclose private and intimate information to others. Some health problems are difficult to be discussed in presence of an interpreter or a member of the family [108]. Therefore, research into working with interprets or health agents and studies of the processes of transmission and counter-transmission are also required.
Shortages in the provision of appropriate training of professional groups could be addressed through the involvement of service users from migrant groups, who could help to overcome the problems of racial and cultural stereotyping which have also been reported in many services.
The absence of monitoring and consultation with clients, which makes impossible for planners and providers to evaluate the suitability of services, are other deficiencies that characterises mental health provision for migrants in Europe.
In order to address all these deficiencies, it is necessary to provide with an adequate financing and a continuity of the grants for research into the multicultural health demand. Most of the projects accomplished have been characterised for a short-term funding.
Finally, there is to highlight the importance of adopting an integrated approach to mental health care that moves away from psychiatric care only, as it has been stressed in a recent report of the WHO in collaboration with Red Cross and Red Crescent organisations.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MGC, MB, MCH and JMHA conceived of the manuscript, and drafted it. All authors read and approved the final manuscript. The "Report on the Mental Health in Europe" working group discussed the manuscript during a meeting.
Acknowledgements
Many thanks to the working group of the "Report on the Mental Health in Europe": Project leader
Viviane Kovess MGEN Foundation of Public Health, Paris 5 University, France
Co-ordinating board
Terry Brugha University of Leicester, UK
Mauro Giovanni Carta University of Cagliari, Italy
Ville Lehtinen STAKES, National Research and Development Centre for Welfare and Health, Finland
Topic experts
Matthias C Angermeyer (Older people) University of Leipzig, Germany
Mariola Bernal (Immigrants) Sant Joan de Deu-SSM, Spain
Miguel Xavier (Substances) Faculty Medical Sciences of Lisbon, Portugal
France Kittel (Gender) ESP ULB Campus Erasme, Belgium
Tom Fryers (Deprivation) University of Leicester, UK
National experts
Bairbre Nic Aongusa Department of Health and Children, Ireland
Claes-Goran Stefansson The National Board of Health and Welfare, Sweden
Henrik Day Poulsen Copenhagen University Hospital, Rigshospitalet, Denmark
Charles Pull Centre Hospitalier de Luxembourg, Luxembourg
Josep Maria Haro Abad Sant Joan de Deu-SSM, Spain
Heinz Katschnig University of Vienna, Austria
Michael G Madianos University of Athens School of Nursing, Greece
Odd Steffen Dalgard University of Oslo, Norway
Rob Bijl Ministry of Justice Research & Documentation, Netherlands
Viviane Kovess MGEN Foundation of Public Health, Paris 5 University, France
Mauro Giovanni Carta University of Cagliari, Italy
Ville Lehtinen STAKES, National Research and Development Centre for Welfare and Health, Finland
Matthias C Angermeyer University of Leipzig, Germany
Miguel Xavier Faculty Medical Sciences of Lisbon, Portugal
France Kittel ESP ULB Campus Erasme, Belgium
Tom Fryers University of Leicester, UK
Other experts
Wolfgang Rutz WHO Regional Office for Europe
John H Henderson Mental Health Europe
Gaetan Lafortune OECD
National referees
Raimundo Mateos Faculty of Medecine Santiago of Compostela, Spain
Paul Bebbington University College London Medical School, UK
José Miguel Caldas de Almeida Pan American Health Organization, Portugal
Alv Dahl Aliv University Hospital, Norway
Matti Joukaama University of Oulu, Finland
Venetsanos Mavrey University of Ionnina Greece
Pierluigi Morosini National Institute of Health, Italy
Per Nettelbladt Lund University, Sweden
Johan Ormel University of Groningen, Netherlands
Frédéric Rouillon University Paris XII, France
Dermot Walsh Health Research Board Dublin, Ireland
Johannes Wancata University of Vienna, Austria
Siegfried Weyerer Organisation Central Institute of Mental Health, Germany
Koen Demyttenaere KU Lueven, Belgium
Managing editor
Karen McColl UK
Other collaborators
Frederic Capuano, Jocelyne Gagnon, Maria Carolina Hardoy, Trevor Hill, Zoe Morgan, Nick Taub, and Jane Smith.
==== Refs
United Nations Number of world's migrants reaches 175 million mark 2002 New York: Press Release POP/844 http://www.un.org
Council of Europe 7th Conference of Ministers Responsible for Migration Affairs: 16–17 September 2002; Helsinki 2002 Council of Europe Bruxelles
Salt Current trends in international migration in Europe 2001 Bruxelles: Council of Europe http://www.mir.es
United Nations Population Division. Replacement Migration Is it a Solution to Declining and Ageing Populations? 2000 New York: UN
Arango J García Castaño FJ, Muriel López C La inmigración en España a comienzos del siglo XXI. Un intento de caracterización La Inmigración en España: Contextos y Alternativas 2002 2 Granada: Laboratorio de Estudios Interculturales
Webster 9th Collegiate Dictionary 1983 Springfield: Merriam-Webster
Watters C Ingleby D Watters C, Ingleby D Introduction to the final report Final Report on Good Practices in Mental Health and Social Care Provision for Asylum Seekers and Refugees 2003 Bruxelles: European Refugee Fund, European Commission
El País Newspaper Madrid 06-08-01/11-11 2003 http://www.elpais.es
Massey D Arango J Hugo G Kouaouci A Pellegrino A Taylor JE Worlds in Motion: Understanding International Migration at the End of the Millenium 1998 Oxford: Claredon Press
Aparicio R García Castaño FJ, Muriel López C La inmigración en el siglo XXI: Las novedades de las actuales migraciones La inmigración en España: contextos y alternativas. Actas del III Congreso sobre la inmigración de España 2002 II Granada: Laboratorio de Estudios Interculturales
Lurbe Puerto K Solé C Incursión sociológica sobre la enajenación de l@s otr@s. Estudio sobre el tratamiento de la diferencia étnica en la salud mental Trabajo de investigación 2002 Barcelona: Universitat Autónoma de Barcelona
Buendía J Estrés y Psicopatología 1993 Madrid: Pirámide
Achotegui J La depresión en los inmigrantes: una perspectiva transcultural 2002 Barcelona: Mayo
Bhui K Abdi A Abdi M Pereira S Dualeh M Robertson D Sathyamoorthy G Ismail H Traumatic events, migration characteristics and psychiatric symptoms among Somali refugees-preliminary communication Social Psychiatry and Psychiatric Epidemiology 2003 38 35 43 10.1007/s00127-003-0596-5
Lie B A 3-year follow-up study of psychosocial functioning and general symptoms in settled refugees Acta Psychiatrica Scandinavica 2002 106 415 25 10.1034/j.1600-0447.2002.01436.x 12392484
LeTouze D Watters C Watters C, Ingleby D Good practices in mental health and social care provision for refugees and asylum seekers. Report on the United Kingdom Final Report on Good practices in mental health and social care provision for asylum seekers and refugees 2003 Bruxelles: European Commission (European Refugee Fund)
Zapata de la Vega B López B Aspectos de la situación legal de los magrebíes en la comarca de L'Horta (Valencia) Inmigración magrebí en España. El retorno de los moriscos 1993 Madrid: Mapfre
Jabardo V López B Inmigrantes magrebíes en la agricultura: La Vega Baja del Segura (Orihuela) Inmigración magrebí en España. El retorno de los moriscos 1993 Madrid: Mapfre
Celaya C López B La vuelta a los desprecios. Algunas consideraciones sobre la xenofobia en la España de los noventa Inmigración magrebí en España. El retorno de los moriscos 1993 Madrid: Mapfre
Bernal M Watters C, Ingleby D Good practices in mental health and social care provision for refugees and asylum seekers. Report on Spain Final report on good practices in mental health and social care provision for asylum seekers and refugees 2003 Bruxelles: European Commission (European Refugee Fund)
Murphy HB Migration culture and mental health Psychological Medicine 1977 7 677 684 594247
Collignon R Immigration et psychopathologie Psychopathologie Africaine 1991 23 265 278
Rack PH Cox JL Migration and Mental Illness Trancultutal Psychiatry 1986 London: Croom Helm
Carta MG Coppo P Reda MA Hardoy MC Carpiniello B Depression and social change. From transcultural psychiatry to a constructivist model Epidemiologia e Psichiatria Sociale 2001 10 46 58 11381479
Odegaard O Migration and insanity. A study of mental disease among the Norwegian population of Minnesota Acta Psych Neurol Scand 1932 4 201 232
Sartorius N Santé méntale et approximation culturelle: le cas du migrant Antropologia Medica 1988 4 8 14
Council of Europe Health conditions of migrants and refugees in Europe Report of the Committee on Migration, Refugees and Demography Doc 8650 2000 Bruxelles: Council of Europe
London MM Mental illness among immigrant minorities in the United Kingdom Brithish Journal of Psychiatry 1986 149 265 73
Cochrane R Bal SS Migration and schizophrenia: an examination of five hypotheses Social Psychiatry 1987 22 181 191 10.1007/BF00583553 3686161
Dean G Walsh D Downing H Shelley E First admissions of native-born and immigrants to psychiatric hospitals in south-east England 1976 British Journal of Psychiatry 1981 139 506 512 7332854
Littlewood R Lipsedge M Psychiatric illness among British Afro-Caribbeans British Medical Journal 1988 296 950 951 3359166
McGovern D Cope RV First psychiatric admission rates of first and second generation Afro-Caribbeans Social Psychiatry 1987 22 139 149 10.1007/BF00583848 3498221
Bhugra D Migration and schizophrenia Acta Psychiatrica Scandinavica 2000 407 68 73 10.1034/j.1600-0447.2000.00015.x
Eaton W Harrison G Ethnic disadvantage and schizophrenia Acta Psychiatrica Scandinavica 2000 407 38 43 10.1034/j.1600-0447.2000.00007.x
Sharpley M Hutchinson G Mc Kenzie K Murray RM Understanding the excess of psychosis among the African-Caribbean population in England. Review of current hypotheses Br J Psychiatry 2001 40 60 68 10.1192/bjp.178.40.s60
Patel SP Gaw AC Suicide among immigrants from the Indian subcontinent: a review Psychiatr Serv 1996 47 517 521 8740494
Karmi G Huismann A, Weilandt C, Greiger Migration and health in UK Reports on Migration and Health in Europe 1997 Bonn: Wissenschaftliches Institut der Ärzte Deutschlands
Jacob KS Bhugra D Lloyd KR Mann AH Common mental disorders, explanatory models and consultation behaviour among Indian women living in the UK J R Soc Med 1998 91 66 71 9602740
Hull SA Cornwell J Harvey C Eldrige S Omo Bare P Prescribing rates for psychotropic medication amongst east London general practices: low rates where Asian population are greatest Family Practice 2001 18 167 173 10.1093/fampra/18.2.167 11264267
Balajaran R Busulu L Adelstein AM Shukla V Patterns of mortality among immigrants to England and Wales from the Indian subcontinent British Medical Journal 1984 289 1185 1187 6437478
Lipsedge M Turner D Services for migrant drug abusers The Lancet 1990 10 1194 10.1016/0140-6736(90)92810-5
Selten JP Cantor-Graee E Slaets J Kahn RS Odegaard's selection hypothesis revisited: schizophrenia in Surinamese immigrants to the Netherlands Am J Psychiatry 2002 159 669 671 10.1176/appi.ajp.159.4.669 11925311
Schrier AC van de Weterin BJ Mulder PJ Selten GP Point prevalence of schizophrenia in immigrants groups in Rotterdam: data from outpatients facilities Eur Psychiatry 2001 16 162 166 10.1016/S0924-9338(01)00558-2 11353594
Selten JP Sijben N First admission rates for schizophrenia in immigrants to the Netherlands. The Dutch National Register Social Psychiatry and Psychiatric Epidemiology 1994 29 71 77
Veen N Selten JP Hoek HW Feller W van der Graaf Y Kahn R Use of illicit substances in a psychosis incidence cohort: a comparison among different ethnic groups in the Netherlands Acta Psychiatrica Scandinavica 2002 105 440 443 10.1034/j.1600-0447.2002.01222.x 12059848
de Jong J The plight and relisience of refugees Lecture at the DTGPP Satellite Conference, Ethnicity & Mental Health in Europe 2003 Essen
Bengi-Arslan L Verhulst FC Crijnen AA Prevalence and determinants of minor psychiatric disorder in Turkish immigrants living in the Netherlands Soc Psychiatry Psychiatr Epidemiol 2002 37 118 124 10.1007/s001270200003 11990008
ten Have ML Bijl RV Inequalities in mental health care and social services utilisation by immigrant women The European Journal of Public Health 2002 9 45 51 10.1093/eurpub/9.1.45
Crijnen AA Bengi Arslan L Verhulst FC Teacher-reported problem behaviour in Turkish immigrant and Dutch children: a cross cultural comparison Acta Psychiatrica Scandinavica 2000 102 439 44 10.1034/j.1600-0447.2000.102006439.x 11142433
Carta MG Carpiniello B Rudas N L'èmigration senegalaise et marocaine en Sardaigne Psychopathologie Africaine 1991 3 329 352
Carta MG Kovess V Hardoy MC Morosini PL Murgia S Carpiniello B Psychiatric disorders in Sardinian immmigrants in Paris: a comparison with Parisians and Sardinians resident in Sardinia Social Psychiatry and Psychiatric Epidemiology 2002 37 112 117 10.1007/s001270200002
Carta MG Kovess V Morosini PL Hardoy MC Marroccu C Carpiniello B Epidemiology study on the prevalence of psychiatric disorders among Sardinian immigrants in Paris compared with Parisians and Sardinians resident in Sardinia Ann Ig 2001 13 139 149 11414103
Emmenegegger T Illness and migration: problems of adjustment of the return migrant Studi Emigr 1988 25 127 136 12341944
Toffoletto D Migranti anziani in Europa 2001 Roma: Psicomedia http://www.psychomedia.it
Carta MG Cetkovich-Bakmans M Hardoy MC Consul ME Carpiniello B Rudas N Depressive disorders in Sardinian emigrants in Argentina: a comparison with Sardinian residents in Sardinia Rendiconti Seminario Facoltà Scienze Università di Cagliari 2003 Suppl 208 211
Hansen C Yagdiran O Mass R Krausz M Potential for misdiagnosis among Turkish migrants with psychotic disorders: a clinical controlled study in Germany Acta Psychiatrica Scandinavica 2000 101 125 129 10.1034/j.1600-0447.2000.90065.x 10706012
Hansen C Yagdiran O Mass R Krausz M Schizophrenic disorders among Turkish migrants in Germany. A controlled study Psychopathology 2001 34 203 208 10.1159/000049308 11549931
Grube M Evaluation of a special project for treatment of psychiatrically ill Turkish migrants Psychiatr Prax 2001 28 81 83 10.1055/s-2001-11580 11305163
Storch G Poutska F Mental disorders in hospitalized children of migrant families from the Mediterranean Prax Kinderpsychol Kinderpsychiatr 2000 49 199 208 10771581
Hjern A Wicks S Dalman C Social adversity contributes to high morbidity in psychoses in immigrants – a national cohort study in two generations of Swedish residents Psychol Med 2004 34 1025 1033 10.1017/S003329170300148X 15554573
Kleiner RI Parker I Social psychological aspects of migration Behaviour in new environments 1979 New York: Sage
Zolkowska K Cantor-Grae E Mc Neil TF Increased rates of psychosis among immigrants to Sweden: is migration a risk factor for psychosis? Psychol Med 2001 31 669 678 10.1017/S0033291701003786 11352369
Baarnhiem S Ekblad S Turkish migrant women encountering health care in Stockolm: a qualitative study of somatization and illness meaning Cult Med Psychiatry 2000 24 431 454 10.1023/A:1005671732703 11128626
Bayard-Burfield L Sundquist J Johansson SE Ethnicity, self reported psychiatric illness, and intake of psychotropic drugs in five ethnic groups in Sweden Epidemiol Community Health 2001 55 675 664
Iglesia E Robertson E Johansonn SE Engfeld P Sundquist J Women, International migration and self-reported health. A population based study of women of reproductive age Soc Sci Med 2003 56 111 24 10.1016/S0277-9536(02)00013-8 12435555
Garrone G Migration and Mental Health in Europe Antropologia Medica 1988 4 23 31
Hjern A Allebeck P Alcohol related disorders in first and second generation immigrants in Sweden: a national cohort study Addiction 2004 99 229 236 10.1046/j.1360-0443.2003.00562.x 14756715
Hjern A Lindbland F Vinnerljung B Suicide, psychiatric illness, and social maladjustment in intercountry adoptees in Sweden: a cohort study The Lancet 2002 360 443 448 10.1016/S0140-6736(02)09674-5
Hoffmann AL Backgard P Beck B Brondum Nielsen K Causes of mental retardation in children of immigrant background. A registry study in the occurrence of consanguinity among parents of mentally retarded children at the Center for Handicapped, Glostrup Hospital, county of Copenhagen Ugeskr Laeger 2002 30 42 46
Esteban y Pena MM Motives for consultation and demographic characteristics of a community of "undocumented" immigrants in the district of Usera-Villaverde (Madrid) Aten Primaria 2001 27 25 28 11218970
Pertinez J Vilades Jene Clusa G Menacho I Nadal S Muns M Descriptive study of mental disorder in ethnic minorities residing in an urban area of Barcelona Aten Primaria 2002 29 6 13
SAPPIR Annual Report. Barcelona 2001
Fossion P Ledoux Y Valente F Servais L Staner L Pelc Minner P Psychiatric disorders and social characteristics among second-generation Moroccan migrants in Belgium: an age and gender controlled study in a psychiatric emergency department Eur Psychiatry 2002 17 443 450 10.1016/S0924-9338(02)00707-1 12504260
Lemoine P Cappadoro R Marie Cardine M La dépression chez les travailleurs maghrébin immigrés Psychopathologie Africane 1981 17 160 170
Pouget R Cirba R Chiariny JF Castelnau D Psychiatric disorders in immigrant French Moslem Ann Med Psychol 1975 2 541 561
Molines C Sapin C Simeoni MC Gentile S Auquier P Perceived health and migration: a new sanitary approach? Rev Epidemiol Sante Publique 2000 48 145 155 10804424
Carlson HM Nielsen EL Ireland: gender, psychological health, and attitudes toward emigration Psychol Rep 1995 76 179 186 7770565
Carta MG Rudas N Psychiatria Transculturale 1998 Cagliari: CUEC
Guidano VF Selfhood processes and life span development 1988 New York: Guilford
Reda MA Sistemi cognitivi complessi e psicoterapia 1986 Roma: La Nuova Italia Scientifica
Csikszentmihalyi The evolving self. A psychology for the third millennium 1993 New York: Harper Collins
Carta MG Carpiniello B Dazzan P Reda MA Depressive symptoms and occupational role among female groups: a research in a south-east african village Psychopathology 2000 32 240 245 10.1159/000029152
Ferron C Haour-Knioe M Tshumper A Narring F Michaud PA Health behaviours adjustment of migrant adolescent in Switzerland Shweiz Med 1997 127 1419 1429
James S Prilleltensky I Cultural diversity and mental health. Towards integrative practice Clin Psychol Rev 2002 22 1133 1154 12436808
Franks F Faux SA Depression stress, mastery and social resources in four ethnocultural women's groups Res Nur Health 1990 13 283 292
Mavreas VR Bebbington PE Greeks, British Greek, Cipriots and Londoners: a comparison of morbidity Psychol Med 1988 18 433 442 3261019
Fitcher MM Elton M Diallina M Koptagel-Ilal G Fthnekis WE Weyerer S Mental illness in Greek and Turkish adolescents Eur Arch Psychiatry Neurol Sci 1988 237 125 134 10.1007/BF00451279 3383918
Aylan S Fisher P Naske R Behavioral problems in children of Turkish guest workers in Vienna Z Kinder Jugendpsychiatr 1993 21 226 232 8147125
Stompe T Bauer S Friedmann A Schizophrenia disorders among migrants DTGPP Satellite Conference, Ethnicity & Mental Health in Europe 2003 Essen
Giannopoulou I Patterns of alcohol and drug use and indicators of psychological distress among migrant greek adolescents Congrès International sur l'alcoolisme et les toxicomanies 1988 2 Oslo: CIAT
Yahyaoui A Toxicomanie et practique sociales 1992 Grenoble: La pensée Sauvage/APPAM
Bendahman H Cultures marginalité et déviance: de l'inquiétante étrangeté de l'autre à l'exclusion sociale 1992 Thionville: Espace Ressource
Yakoub S Réflexions à propos de toxicomanies maghrébins incarcérés: approches juridique, socio-ethnique, ethno-clinique 1993 Paris: Interventions
Boylan M Acculturation et conduites addictives chez les jeunes d'origine maghrébine 1995 Paris: Interventions
Akbiyik O Un centre d'accueil pour toxicomanes migrants 1990 Paris: Bull. Liaison CNDT
Caballero L Addiction disorders among migrants Lecture at the DTGPP Satellite Conference, Ethnicity & Mental Health in Europe 2003 Essen
Carey-Wood J Duke K The settlement of refugees in Britain: Home Office Research Study 141 1995 London: Home Office
Brent and Harrow Health Agency, Brent, Harrow Refugee Survey 1995 London: Northwest London Training and Enterprise Council
Mollicca RF McInnes K Sarajlic N Lavelle J Sarajlic I Massagli M Disability associated with psychiatric comorbidity and health status in Bosnian refugees living in Croatia JAMA 1990 282 433 439 10.1001/jama.282.5.433
Mollica RF Sarajlic N Chernoff M Lavelle J Vukovic IS Massagli MP Longitudinal study of psychiatric symptoms, disability, mortality, and emigration among Bosnian refugees JAMA 2001 286 5 546 554 10.1001/jama.286.5.546 11476656
The State of the Mental Health in Europe 2004 Luxemburg: Directorete Generale Health and Consumer ProtectionEuropean Commission
Silove D Steel Z Watters C Policies of deterrence and the mental health of asylum seekers Journal of the American Medical Association 2000 284 604 611 10.1001/jama.284.5.604 10918707
Jurado Guerrero T Easy Scapegoats: sans papier immigrants in Europe. State strategies and intervention strategies for the civil society Conclusions of a project organised within the DG "Employment and Social Affairs" programme "Preparatory Actions to Combat and Prevent Social Exclusion" 2000 http://www.freudenbergstiftung.de
Watters C Migration and Mental Health care in Europe: report of a preliminary mapping exercise Journal of Ethnic and Migration Studies 2002 28 102 107 10.1080/13691830120103985
Machleidt W Psychosocial issue in the treatment of migrants Lecture at the DTGPP satellite conference, ethnicity & mental health in Europe 2003 Essen
Fernando S Cultural diversity, mental health and psychiatry The struggle against racism 2003 London: Brunner-Routledge
Ete E Koch E, Ozek M, Pfeiffer WM Ethnomedizinische Aspekte der Interaktion mit turkischen Patienten Psychologie und Pathologie der Migration 1995 Freiburg: Lambertus S
Health for all, all in health: Report phase I. "Preparatory measures aimed at combating and preventing discrimination". Community Action Programme 2002. Building an action group against discrimination by health care services Preliminary report 2002 Non-published
|
16135246
|
PMC1236945
|
CC BY
|
2021-01-04 18:01:01
|
no
|
Clin Pract Epidemiol Ment Health. 2005 Aug 31; 1:13
|
utf-8
|
Clin Pract Epidemiol Ment Health
| 2,005 |
10.1186/1745-0179-1-13
|
oa_comm
|
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-151616475610.1186/1745-0179-1-15Case reportSudden psychotic episode probably due to meningoencephalitis and Chlamydia pneumoniae acute infection Xavier Miguel [email protected] Bernardo [email protected] Marta [email protected] Nuno [email protected]ães João [email protected] Depart. Psychiatry and Mental Health, Faculty Medical Sciences – UNL Calçada da Tapada, 155, 1300-Lisbon, Portugal2 Depart. Psychiatry – Hospital S. Francisco Xavier, 1400-Lisbon, Portugal3 Depart. Neurology, Hospital Egas Moniz, 1400-Lisbon, Portugal2005 15 9 2005 1 15 15 6 9 2005 15 9 2005 Copyright ©2005 Xavier et al; licensee BioMed Central Ltd.2005Xavier 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
Since 9% to 20% of all cases of acute psychosis presenting to an Emergency Department (ED) are due to a general medical condition, cautious medical workup should be mandatory in such patients. Differential diagnosis must consider conditions as diverse as renal failure or CNS infection.
Acute Chlamydia pneumoniae infection usually causes a self-limited respiratory syndrome. Rarely, acute neurological complications occur, with acute meningoencephalitis most frequently reported. Diagnosis requires a high level of suspicion and is difficult to confirm.
Case report
We describe a 22 year-old female Caucasian who, three days after a mild pharingitis, developed an acute psychosis with exuberant symptoms interspersed with periods of lucidity, in a background of normal consciousness and orientation.
Initial medical and imagiological workup were inconclusive. After 20 days of unsuccessful treatment with antipsychotics she developed a high fever and was re-evaluated medically. Lumbar puncture revealed an inflammatory cerebrospinal fluid. MRI showed irregular thickening and nodularity of the lateral ventricles' lining. An anti-Chlamydia pneumoniae IgM antibody titter of 85 IU/ml was detected. All symptoms cleared after treatment with antibiotics and corticosteroids.
Conclusion
This is, to our knowledge, the first reported case of acute CP-associated meningoencephalitis manifesting as an acute psychotic episode. It illustrates the principle that non-organic psychiatric syndromes must remain a diagnosis of exclusion in first-time acute psychosis.
==== Body
Background
It is estimated that between 9% and 20% of all cases of acute psychosis presenting to an Emergency Department (ED) are due to a general medical condition [1-3]. This means that all patients coming to the ED with signs and symptoms of acute psychosis must undergo a careful diagnostic workup. Apart from non-organic psychiatric syndromes themselves, the differential diagnosis must include physical trauma, drugs and toxins, organ failure (e.g., renal failure), structural lesions like intracranial hematomas or neoplasms, infections and nutritional deficiencies like vitamin B12 deficiency or pellagra [1-4]. The psychiatrist plays a fundamental role in this process and is often the last resort that keeps the patient from sliding down the almost no-return path of an erroneous non-organic psychiatric diagnosis [2,4,5]. The sudden onset of psychosis in a 40 year-old patient with no personal or family history of psychiatric disease who shows recent memory loss, altered vital signs and a clouded consciousness with disorientation and visual hallucinations will probably be easily identified as a possible medical condition manifesting through behavioural changes [5]. Many cases, however, do not present such clear-cut features and false beliefs (i.e. delusions) may actually constitute the sole manifestation of central nervous system disfunction (CNS) [6]. Such patients often come to constitute a true challenge to the attending psychiatrist.
Chlamydia pneumoniae (CP) is an intracellular organism with worldwide distribution and is pathogenic both to humans and other vertebrates. The metabolically inactive but infectious extracellular form (elementary body) differentiates into reticulate body after endocytosis by macrophages/monocytes, endothelial cells or vascular smooth muscle cells. Within these cells it then replicates by binary fission [7]. It is currently speculated that the respiratory tract be the site of entry for CP, from whence it is then transported by monocytes or macrophages to other sites in the organism. It remains unclear how CP infection leads to disease, as it does not produce toxins and has only weak lipopolysaccharide activity [8]. In humans CP most frequently causes a self-limited and uneventful acute respiratory tract syndrome. More rarely it can cause multiorganic disease with an occasionally fatal outcome [8,7,9]. Over the last few years CP has become the object of intense debate among neurologists as being possibly involved in the aetiology of such diverse conditions as Alzheimer's disease, multiple sclerosis or atherosclerosis (this last possibility earning it the ironical nickname "cardiology's Helicobacter") [10-18]. Acute neurological complications of CP infection seem to be rare [10]; as to our knowledge no more than twenty putative cases with variable manifestations have been published so far [19-25]. Meningoencephalitis has been by far the most frequently reported situation, with isolated reports of meningoradiculitis, polyradiculoencephalitis, cerebelar disfunction, Guillain-Barré syndrome and acute disseminated encephalomyelitis. In most such cases direct evidence of central nervous system (CNS) involvement by CP in has been difficult to obtain; to our knowledge only three cases have been reported where intrathecal production of specific anti-CP antibodies was demonstrated [19-21]. In a further case the authors claim to have detected CP antigens in their patient's CSF and more recently a case was reported where intrathecal presence of CP DNA was detected by PCR [23,25].
Case report
A 22 year-old single, Caucasian female Business-school student developed a low fever, headache and sore throat, which remitted spontaneously after three days. Seven days after the start of these symptoms she suddenly became agitated, physically and verbally aggressive, with total insomnia and disrupted behaviour. She was taken by her parents to the Emergency Department of S. Francisco Xavier Hospital, Lisbon, and was compulsorily hospitalised in the acute psychiatric ward under the Portuguese Mental Health Law. The mental state examination revealed a vigil and fully oriented patient. She appeared distrustful, collaborated poorly and was easily distracted. She had false delusional memories and paranoid delusional, poorly structured ideas of persecutory, magical and grandiose content, as well as a disrupted awareness of the boundaries and vitality of the Self. She also claimed, in what appeared to be a delusional misidentification of the Capgras syndrome type, that robots had replaced her parents. She appeared to be suffering complex auditive hallucinations in the 2nd and 3rd person, with a menacing content. Her mood was dysphoric, with poorly contained, labile affects. Her past medical record and family history were irrelevant. Neurological and general physical examinations were normal. She was started on haloperidol 10 mg tid ev (she refused oral medication), with no improvement of her mental state, although she did have unpredictable, short-lived intervals of remarkable lucidity with almost total absence of psychiatric symptoms apart from a depressed mood. 20 days after admission she developed a high fever with a white-cell count of 27 × 109/L, an erythrocyte sedimentation rate (ESR) of 100 mm/h and a C-reactive protein (CRP) of 15.3 mg/L. Routine biochemistry studies were normal. The cerebrospinal fluid (CSF) was clear, with normal protein and glucose concentrations and 20 × 106 lymphocytes per litre. Microbiological investigation of CSF was negative. Her chest film and brain CT-scan showed no changes. Brain MRI showed irregular thickening and nodularity of the lateral ventricles' lining. A complete neuropsychological evaluation was performed revealing a mild impairment of her external visual attention and visuo-constructive reproduction capacity. The EEG displayed symmetrical, irregular, high amplitude, slow wave activity with frontotemporal predominance. An ophthalmologic examination revealed lymphocytes in her vitreum. A skin biopsy was taken for Lupus Band Test, which was negative. Serologic testing of serum and CSF was performed for systemic lupus erythematosus (including anti-ribosomal P antibodies) and other connectivites as well as for most known neurotropic viruses and bacteria. An IgG antibody titer to CP of 22 IU/ml was detected in the serum, with an IgM antibody titer of 85 IU/ml. No anti-CP antibodies were detected in the CSF. All her psychotropic medication was stopped and she was started on levofloxacin (due to dysuria and leukocyturia) and acyclovir. Five days levofloxacin was changed to gentamicin plus piperacillin/tazobactam for due to Escherichia coli-positive blood-cultures (serological results were not yet available at that time). She finally received three consecutive ev pulses of methylprednisolone (1 g each), before being transferred to a Neurology unit. During the following week she progressively became apyretic and her psychotic symptoms gradually remitted. At the time of discharge, eight weeks after admission, she still had an ESR of 55 mm/h and a CRP of 9.03 mg/L, with normal blood count and biochemical parameters as well as a normal CSF. Brain MRI nine weeks after admission showed complete remission of the initial changes. At the time of discharge, the EEG still displayed symmetrical, irregular, high amplitude, slow wave activity with frontotemporal predominance, although much less severe than in the first EEG record.
Psychiatrically she remained slightly disinhibited, both verbally and affectively, with pedolalia and mildly childish behaviour. There were no signs of delusional ideas or hallucinations and she could only vaguely remember the period when she had been psychotic. Two months after discharge she had resumed her studies and concluded her Business School degree.
Discussion
This is, to our knowledge, the first reported case of an acute CP-associated meningoencephalitis manifesting as a first-episode acute psychosis.
In our patient the initial manifestations were interpreted as a purely non-organic psychiatric syndrome, which led to a delay of several weeks not only in the definitive diagnosis but also in the appropriate treatment. The patient's young age, her clear consciousness and normal orientation, the presence of archetypical first-rank symptoms of schizophrenia and specially her normal physical and neurological examinations were all factors contributing to the premature exclusion of a medical condition as an explanation for the patient's behavioural changes. This illustrates how easy it can be to miss an organic brain syndrome at an early stage once a psychiatric diagnosis is taken into consideration and in the absence of such gross indicators as age greater than 40 years, disorientation with clouded consciousness or abnormal vital signs. Retrospectively we can identify several subtle cues in our patient pointing to a medical aetiology, namely the sudden onset of severe and exuberant psychiatric changes, the patient's high level of functioning prior to the episode and negative family history for psychiatric disorders, her intense emotional unsteadiness, the temporal relationship to a minor infectious episode, the existence of occasional "islands" of lucidity and the symptoms' minimal response to antipsychotic agents. All of these have been consistently reported in the literature as discrete but highly significant indicators of acute organic brain disorder [1,2,4,5,26,27].
A pertinent question in this context is which diagnostic methods should be included in the initial workup of first-episode acute psychosis. Although most would probably include a no-contrast brain CT scan in their initial evaluation, this is actually a method lacking in sensitivity [1-3,27]. On the other hand, MRI scanning is probably unavailable in most EDs. Also, the decision threshold for performing a lumbar puncture should be much lower, since it remains the most reliable method of detecting inflammatory changes of the CNS [1,3].
Our case is also worthy of discussion from the perspective of infectiology, especially insofar as the final bacteriological diagnosis is concerned. CP infection of the CNS is difficult to diagnose and requires a high level of suspicion. Detection of the organism's DNA in the CSF using polymerase chain reaction (PCR) probably constitutes the most reliable method of confirming the diagnosis [25,28,29]. Unfortunately, this is a method that is not always available, as it is the case in our hospital. However, the elevated titer of CP IgM antibodies in the serum and the absence of any other possible aetiology after exhaustive investigation make CP the almost certain cause of meningoencephalitis in our patient. To our knowledge only three cases where intrathecal synthesis of CP antibodies was demonstrated have been published so far and most reports of the rare acute neurological complications of CP infection have been based on serum antibody titers only [19-21]. Moreover, Sočan et al, using a direct-immunfluorescence test with CP-specific monoclonal antibodies, reported a case where they detected CP antigens in their patient's CSF, which showed negative antibody titers by microimmunofluorescence assay [23].
Doubt has been cast by several authors on the reliability of single antibody titer measurements in the serum for the diagnosis of acute CP infection and it now seems unanimous that this method's sensitivity is unsatisfactory [28,29]. The case for specificity, however, remains open to controversy. Although Gaydos et al detected antibody titles considered to be diagnostic of acute infection with CP in 18,8% of 80 PCR and culture negative individuals, it must be stressed that this was a sample of asymptomatic persons and so their results can in no way be extrapolated to severely ill patients presenting with clinical manifestations compatible with CP infection and in whom most alternative infectious agents have been exhaustively excluded. On the other hand, Sočan et al, in their reply to the letter to their editor by Weiss et al, claim to have found no single positive anti-CP IgM titer in samples from 100 blood donors [28]. Finally, it is probably also relevant that the abovementioned works have used microimmunofluorescence assays for measuring antibody titers in the serum, while our patient was studied using an enzyme-linked imunosorbent assay. The whole controversy around which is the most reliable method for detecting CP infection of the CNS actually developed in connection with attempts to link the organism to the aetiology of Multiple sclerosis, Alzheimer's disease or atherosclerosis, leading to the active search for CP in patients with no other evidence of infection by CP than the above-mentioned conditions themselves [11-13,15-17,30]. To be conclusive, this kind of investigation must strive for a degree of certainty and unambiguity in their methods whose sophistication and complexity are probably inadequate for everyday use in a clinical context.
Another important question in our case is what caused the patient to improve. The fact that she only received levofloxacin for five days (serological results were only available after the patient had started to improve) and her rapid recovery after corticosteroid therapy suggests that the acute neurological complications of CP infection might be attributed to an autoimmune mechanism, rather than to the organism's direct action. Frydén et al have previously reported improvement after corticosteroid therapy in CP-associated meningoencephalitis (although their patient also received cloramphenicol) [31]. The CP-associated Guillain-Barré syndrome reported by Haidl also recovered only after treatment with methylprednisolone [24]. The encephalitis case reported by Airas et al suffered further clinical and imagiological progression in spite of CP-active antibiotic treatment (levofloxacin) [19]. The case reported by Michel et al eventually recovered from lumbosacral meningoradiculitis without antibiotic treatment [21]. Finally, Grayston et al described a case where acute CP reinfection led to a sarcoidosis-like clinical picture and raise the possibility of an immunopathologic reaction in their patient [9].
In spite of our current ability to detect structural and functional changes of the CNS in organic brain syndromes, it remains unclear how these changes relate to the behavioural changes and in what way such observations might be extrapolated to our understanding of the pathophysiology of non-organic psychosis. In the present case the only observable changes seemed to concentrate around the ependymal axis. Peri-aqueductal structures have recently been considered by a few authors to be the neurostructural backbone of consciousness and especially of the consciousness of the self, whose disturbance constitutes one of essential features of schizophrenic psychosis [32,33]. This concept is in full accordance to the observations by Cummings et al, who found that in organic psychosis complex symptomatology such as Schneiderian first-rank symptoms is more frequently associated with subcortical lesions. This probably means that complex psychotic syndromes require that higher CNS functions like linguistic and verbally-mediated conceptual abilities and the corresponding cortical structures be intact. Subcortically generated abnormal emotional experiences can thus be fully elaborated at higher levels into complex delusional and hallucinatory syndromes [6]. Interestingly, organic psychosis due to subcortical brain lesions also seem to be particularly resistant to antipsychotic treatment, which was one of the striking features of our patient [6].
Conclusion
As we stressed before, this is to our best knowledge the first reported case of an acute CP-associated meningoencephalitis manifesting as a first-episode acute psychosis and illustrates the overwhelming importance of a careful medical and neurological diagnostic workup before a sudden first-time psychotic episode receives a "non-organic" psychiatric diagnosis. In such cases non-organic psychiatric diagnosis should remain a diagnosis of exclusion [1,4], and thus approached with the assumption that the cause of psychosis is an organic one [4]. As diagnostic methods grow more complex and as we become more able to diagnose unusual medical aetiologies for common psychiatric syndromes, more and more will be demanded from clinical psychiatrists in terms of general medical knowledge and ability to work fully integrated with other specialities.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MX wrote and submitted the manuscript
BC helped to draft the manuscript and reviewed published references
MC and NC have made substantial contributions to acquisition and interpretation of clinical data:
JG revised the manuscript critically from the neurological point of view
All authors read and approved the final manuscript. Moreover, all authors were involved in the care of the patient described in this case report
==== Refs
Richards CF Gurr DE Psychosis Emerg Med Clin North Am 2000 18 2 253 262 10.1016/S0733-8627(05)70122-X 10767882
Duwe BV Turetsky BI Misdiagnosis of Schizophrenia in a Patient With Psychotic Symptoms Neuropsychiatry, Neuropsychol Behav Neurol 2002 15 252 260
Talbot-Stern JK Green T Royle TJ Psychiatric Manifestations of Systemic Illness Emerg Med Clin North Am 2000 18 2 199 209 10.1016/S0733-8627(05)70118-8 10767878
Frame DS Kercher EE Acute Psychosis: Functional Versus Organic Emerg Med Clin North Am 1991 9 1 123 136 2001662
Dubin WR Weiss KJ Zeccardi JA Organic Brain Syndrome: The Psychiatric Impostor JAMA 1983 249 60 62 10.1001/jama.249.1.60 6848782
Cummings JL Organic Delusions: Phenomenology, Anatomical Correlations, and Review Br J Psychiatry 1985 146 184 197 3156653
Chirgwin K Roblin PM Gelling M Hammerschlag MR Schachter J Infection with Chlamydia pneumoniae in Brooklin J Infect Dis 1991 163 757 761 2010629
Kuo C-C Jackson LA Campbell LA Grayston JT Chlamydia pneumoniae (TWAR) Clin Microbiol Rev 1995 8 4 451 461 8665464
Grayston JT Campbell LA Kuo C-C Mordhorst CH Saikku P Thom DH Wang S-P A New Respiratory Tract Pathogen: Chlamydia pneumoniae Strain TWAR J Infect Dis 1990 161 618 625 2181028
Yucesan C Sriram S Chlamydia pneumoniae infection of the central nervous system Curr Opin Neurol 2001 14 355 359 10.1097/00019052-200106000-00015 11371760
Gaydos CA Chlamydia pneumoniae and its link to multiple sclerosis: To be or not to be? Neurology 2001 56 1126 1127 11342673
Jacobson S Cross A Association of Chlamydia pneumoniae and multiple sclerosis: Stage two? Neurology 2001 56 1128 1129 11342674
Griggs RC Chlamydia: conflict and controversy Neurology 2001 56 1130 11342675
Woessner R Treib J Chlamydieninfektionen in der Neurologie Dtsch Med Wschr 2001 126 153 155 11233885
Boman J Roblin PM Sundström P Sandström M Hammerschlag MR Failure to detect Chlamydia pneumoniae in the central nervous system of patients with MS Neurology 2000 54 265 10636169
Sriram S Failure to detect Chlamydia pneumoniae in the central nervous system of patients with MS Neurology 2000 55 1 1423 1424 Letter 11087807
Numazaki K Chibar S Failure to detect Chlamydia pneumoniae in the central nervous system of patients with MS Neurology 2001 57 2 746 Letter 11524506
Hunter S Hafler A Ubiquitous pathogens – Links between infection and autoimmunity in MS? Neurology 2000 55 2 164 165 10908883
Airas L Kotilainen P Vainionpää R Marttila RJ Encephalitis associated with Chlamydia pneumoniae Neurology 2001 56 1778 1779 11425958
Guglielminotti J Lallouche N Maury E Alzieu M Guidet B Offenstadt G Severe Meningoencephalitis: An Unusual Manifestation of Chlamydia pneumoniae Infection Clin Infect Dis 2000 30 209 210 10.1086/313633 10619758
Michel D Antoine JC Pozzetto B Gaudin OG Lucht F Lumbosacral meningoradiculitis associated with Chlamydia pneumoniae infection J Neurol Neurosurg Psychiatry 1992 55 511 1619428
Korman TM Turnidge JD Grayson ML Neurological Complications of Chlamydial Infections: Case Report and Review Clin Infect Dis 1997 25 847 851 9356800
Sočan M Beovič B Keše D Chlamydia pneumoniae and meningoencephalitis N Engl J Med 1994 331 406 10.1056/NEJM199408113310618
Haidl S Ivarsson S Bjerre I Persson K Guillain-Barré syndrome after Chlamydia pneumoniae infection N Engl J Med 1992 326 576 577 1732807
Minnerop M Bös Harbrecht U Maass M Urbach H Klockgether T Schröder R CNS infection with Chlamydia pneumoniae complicated by multiple strokes J Neurol 2002 249 1329 1331 10.1007/s00415-002-0833-y 12242570
Khan S Haddad P Montague L Summerton C Systemic lupus erythematosus presenting as mania Acta Psychiatr Scand 2000 101 406 408 10.1034/j.1600-0447.2000.101005406.x 10823302
Muñoz-Málaga A Anglada JC Páez M Girón JM Barrera A Psicosis como manifestación inicial de lupus eritematoso sistémico: valor de la prueba de la banda lúpica frente a los anticuerpos antirribosomales Rev Neurol 1999 28 8 779 781 10363321
Weiss SM Roblin PM Hammerschlag MR More on Chlamydia pneumonia and meningoencephalitis N Engl J Med 1995 332 62 63 Letter 10.1056/NEJM199501053320115 7990880
Gaydos CA Roblin PM Hammerschlag MR Hyman CL Eiden JJ Schachter J Quinn TC Diagnostic Utility of PCR-Enzyme Imunoassay, Culture, and Serology for Detection of Chlamydia pneumoniae in Symptomatic and Asymptomatic Patients J Clin Microbiol 1994 32 4 903 905 8027341
Yao S-Y Stratton CW Mitchell WM Sriram S CSF oligoclonal bands in MS include antibodies against Chlamydophila antigens Neurology 2001 56 1168 1176 11342681
Frydén A Kihlström E Maller R Persson K Romanus V Ånséhn S A clinical and epidemiological study of ornithosis caused by Chlamydia psitacci and Chlamydia pneumoniae (strain TWAR) Scand J Infect 1989 21 681 691
Panksepp J Affective Neuroscience: The Foundations of Human and Animal Emotions 1998 Oxford University Press, New York and Oxford
Damasio AR Thomas Metzinger A Neurobiology for Consciousness Neural Correlates of Consciousness 2000 The MIT Press, Cambridge (Massachussets) and London 25 30
|
16164756
|
PMC1236946
|
CC BY
|
2021-01-04 18:01:01
|
no
|
Clin Pract Epidemiol Ment Health. 2005 Sep 15; 1:15
|
utf-8
|
Clin Pract Epidemiol Ment Health
| 2,005 |
10.1186/1745-0179-1-15
|
oa_comm
|
==== Front
Curr Control Trials Cardiovasc MedCurrent Controlled Trials in Cardiovascular Medicine1468-67081468-6694BioMed Central 1468-6708-6-121611532110.1186/1468-6708-6-12ResearchEnzyme estimates of infarct size correlate with functional and clinical outcomes in the setting of ST-segment elevation myocardial infarction Turer Aslan T [email protected] Kenneth W [email protected] Dianne [email protected] W Douglas [email protected] Robert H [email protected] Nathan R [email protected] E Magnus [email protected] Department of Internal Medicine, Duke University Medical Center and Duke Clinical Research Institute, Durham, North Carolina, USA2 Henry Ford Hospital, Detroit, Michigan, USA3 University of Maryland, Baltimore, Maryland, USA4 University of Washington, Seattle, Washington, USA5 University of North Carolina, Chapel Hill, North Carolina, USA2005 23 8 2005 6 1 12 12 29 7 2005 23 8 2005 Copyright © 2005 Turer et al; licensee BioMed Central Ltd.2005Turer 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
Cardiac biomarkers are routinely obtained in the setting of suspected myocardial ischemia and infarction. Evidence suggests these markers may correlate with functional and clinical outcomes, but the strength of this correlation is unclear. The relationship between enzyme measures of myocardial necrosis and left ventricular performance and adverse clinical outcomes were explored.
Methods
Creatine kinase (CK) and CK-MB data were analyzed, as were left ventricular ejection fraction (LVEF) by angiogram, and infarct size by single-photon emission computed tomography (SPECT) imaging in patients in 2 trials: Prompt Reperfusion In Myocardial-infarction Evolution (PRIME), and Efegatran and Streptokinase to Canalize Arteries Like Accelerated Tissue plasminogen activator (ESCALAT). Both trials evaluated efegatran combined with thrombolysis for treating acute ST-segment elevation myocardial infarction (STEMI).
Results
Peak CK and CK area-under-the-curve (AUC) correlated significantly with SPECT-determined infarct size 5 to 10 days after enrollment. Peak CK had a statistically significant correlation with LVEF, but CK-AUC and LVEF correlation were less robust. Statistically significant correlations exist between SPECT-determined infarct size and peak CK-MB and CK-MB AUC. However, there was no correlation with LVEF for peak CK-MB and CK-MB AUC. The combined outcome of congestive heart failure and death were significantly associated with CK AUC, CK-MB AUC, peak CK, and peak CK-MB measurements.
Conclusion
Peak CK and CK-MB values and AUC calculations have significant correlation with functional outcomes (LVEF- and SPECT-determined infarct size) and death or CHF outcomes in the setting of STEMI. Cardiac biomarkers provide prognostic information and may serve as valid endpoint measurements for phase II clinical trials.
==== Body
Background
There has been considerable interest in validating established cardiac biomarkers as a mechanism for assessing infarct size in clinical practice and as an endpoint in clinical trials. Since assays for creatine kinase (CK) and CK-MB became widely available in the 1970s, attempts have been made to correlate the levels of these biomarkers with outcomes and infarct size using kinetic models and curve-fitting techniques [1-3]. Despite initial enthusiasm, these models were later criticized because of their inability to accurately predict the extent of myocardial necrosis [4-8].
Studies in animals [4,6,9,10] and humans [5,11,12] have shown a significant correlation between CK and CK-MB-derived estimations of myocardial damage after acute myocardial infarction (MI) and the extent of damage at necropsy, albeit in the era prior to reperfusion therapy. Research has suggested that the quantity of cardiac markers released correlates with infarct size and with clinical outcomes such as arrhythmias [13,14], heart failure [13,15-17], and mortality [13,16,18,19] in the setting of both ST-segment elevation myocardial infarction (STEMI) and non-STEMI (NSTEMI). Since cardiac biomarkers are relatively inexpensive to measure and are routinely used in clinical practice, they are an attractive tool for easily determining infarct size and gathering prognostic information.
Little has been done to determine how the different methods used to assess infarct size in acute MI patients treated with reperfusion therapy correlate. Improving our understanding of the methods used to measure infarct size and clarifying the association between infarct size and outcomes can be of tremendous value to clinical trials, particularly in the early evaluation of new therapies or interventions.
The aims of the current analyses were to 1) determine whether enzyme-determined infarct size would correlate to the degree of infarction measured by single-photon emission computed tomography (SPECT) imaging and to left ventricular ejection fraction (LVEF) measured by LV angiography, and 2) assess whether the amount of enzyme released could be used to predict clinical outcomes in patients treated with thrombolysis. For this analysis, data from the Efegatran and Streptokinase to Canalize Arteries Like Accelerated Tissue plasminogen activator (ESCALAT) [20] and Prompt Reperfusion In Myocardial-infarction Evolution (PRIME) [21] trials were used.
Methods
The trials
The ESCALAT trial has been published [20]. In brief, ESCALAT was a randomized, dose-finding study in which intravenous (IV) efegatran sulfate (in 1 of 4 doses) plus streptokinase, or IV heparin plus accelerated tissue plasminogen activator (t-PA) was given to patients (N = 245) presenting with acute STEMI. Trial participants were between 21 and 75 years old, had ischemic chest pain for ≥ 30 minutes associated with ST-segment elevation of ≥ 0.1 mV in 2 or more contiguous electrocardiographic leads, and onset of symptoms within 12 hours of planned treatment. The primary endpoint of infarct-related artery patency was assessed by angiography at 90 minutes.
In the PRIME trial [21], patients (N = 336) with acute MI were randomly assigned to accelerated t-PA and either IV efegatran sulfate (in 1 of 5 doses) or IV heparin. In PRIME, inclusion and exclusion criteria were similar to those for ESCALAT [20,21]. The study protocol required coronary angiography 90 minutes after start of t-PA therapy and repeat angiography 5 to 7 days later.
SPECT and left ventricular (LV) functional assessments
SPECT quantification of infarct size was performed 5 to 10 days after randomization per the ESCALAT protocol. SPECT images were evaluated in blinded fashion by a central core laboratory (Georgetown University, Washington, DC) using previously described methodologies [22].
In the PRIME trial, patients who did not have percutaneous coronary intervention or bypass surgery were to have repeat coronary angiography 5 to 7 days after enrollment or prior to hospital discharge, whichever occurred first. All angiograms were reviewed by an independent evaluator, blinded to treatment assignment, at a central core laboratory (Cleveland Clinic, Cleveland, Ohio). Angiograms were assessed for vessel patency, Thrombolysis In Myocardial Infarction (TIMI) flow grades, and LV function.
Biomarker determination of infarct size
In both the ESCALAT and PRIME trials, CK and CK-MB levels were to be drawn at baseline (time of enrollment) and at 6, 12, 24, 36, and 72 hours after enrollment. All samples were analyzed at a central core laboratory (University of Maryland, College Park, MD), and the evaluator was blinded to patient information. Cardiac biomarker curves were then plot-fitted by the method proposed by Vollmer and colleagues [23]. Area-under-the-curve (AUC) and peak biomarker levels were derived for both CK and CK-MB from these curve-fit data (Figure 1).
Figure 1 Log normal plot of total CK data for a typical patient. The connected points are from curve fitting, the solid circles are the actual measured points, and the solid triangles (along x-axis) are the residuals. In this case, the measured 8-hour point correlates well with the calculated peak for the curve. CK: Creatine kinase, TCK: Total CK.
Clinical outcomes
Clinical outcomes of death or congestive heart failure (CHF) that occurred during the index hospitalization were reported by the investigator with standard data collection tools without adjudication using standard definitions.
Statistical analysis
Statistical analyses were performed with SAS software (version 6, SAS, Inc., Cary, North Carolina) on the UNIX system. Continuous variables were summarized using medians and interquartile ranges whereas categorical variables were summarized as frequencies and percentages.
Patients from both trials were excluded from analyses if they had a history of MI or no elevation of CK or CK-MB due to confounding. Within the ESCALAT trial, after these exclusions, the study size was 136 patients, representing 56% of the original study population. From the PRIME trial, 165 patients were included in the analysis after exclusions, accounting for 49% of the original trial population (Figure 2).
Figure 2 PRIME and ESCALAT study populations. CK: Creatine kinase, MI: Myocardial infarction. SPECT, Single-photon emission computed tomography.
Spearman rank correlations were used to determine associations between the cardiac biomarker data and size of infarct as measured by SPECT for the ESCALAT trial and LVEF for the PRIME trial. Wilcoxon rank sum tests were used to determine differences in continuous variables between PRIME and ESCALAT.
Results
Baseline demographics
The baseline demographics were similar for the ESCALAT and PRIME trial populations with respect to age, sex, race, and history of smoking. None of the patients had a history of MI as part of the inclusion criteria for this analysis.
Clinical outcomes
There were no significant differences in death or CHF during the index hospitalization between the treatment groups in either trial. There were also no statistical differences between the trials with respect to CK and CK-MB data (Table 1); therefore, the trials were combined for further analysis of enzyme release. Table 1 shows the infarct size assessments by peak CK, peak CK-MB, CK-AUC, CK-MB AUC, LVEF, and SPECT data.
Table 1 ESCALAT and PRIME trials: assessments of myocardial damage
ESCALAT n = 136 PRIME n = 165 P value
Peak CK
n 127 158
Median (25th, 75th) 1790.0 (682, 2580) 1730.0 (931, 2850) 0.6718
Peak CK-MB
n 129 161
Median (25th, 75th) 209 (82, 340) 207 (101, 362) 0.6239
CK-MB AUC
n 129 161
Median (25th, 75th) 3887.0 (1617, 6099) 3891.0 (2035, 6746) 0.1588
CK AUC
n 127 158
Median (25th, 75th) 50960 (22950, 79620) 51295 (29930, 87390) 0.1801
LVEF (%)
n n/a 52
Median (25th, 75th) n/a 58.0 (50, 66) n/a
SPECT (% infarcted)
n 122 n/a
Median (25th, 75th) 13.0 (4, 24) n/a n/a
P values obtained using Wilcoxon rank sum test.
AUC: Area-under-the-curve, CK: creatine kinase, ESCALAT: Efegatran and Streptokinase to Canalize Arteries Like Accelerated Tissue plasminogen activator, LVEF: left ventricular ejection fraction, PRIME: Prompt Reperfusion In Myocardial infarction Evolution, SPECT: single-photon emission computed tomography
Correlation of enzymes and SPECT
In ESCALAT, no differences in infarct size by SPECT imaging, CK-AUC, or CK-MB AUC were observed between the efegatran-plus-streptokinase- and heparin plus t-PA-treated groups. Therefore, all treatment groups were combined for these analyses. The median (25th, 75th) time to SPECT was 6 days [5,16].
A scatter-plot of CK-MB AUC and SPECT infarct size is shown in Figure 3. Statistically significant positive correlations between the SPECT-determined infarct size and the peak values for CK (r = 0.65, P < 0.0001) and CK-MB (r = 0.64, P < 0.0001) were observed. SPECT infarct size also correlated with the AUC measurements for both CK (r = 0.63, P < 0.0001) and CK-MB (r = 0.58, P < 0.0001) (Table 2).
Table 2 ESCALAT and PRIME trials: correlations of enzymes with LVEF- and SPECT-determined infarct size
Peak CK Peak CK-MB CK AUC CK-MB AUC
LVEF r = -0.30 r = -0.26 r = -0.28 r = -0.22
P = 0.0354 P = 0.0599 P = 0.0508 P = 0.1117
n = 50 n = 51 n = 50 n = 51
SPECT r = 0.65 r = 0.64 r = 0.63 r = 0.58
P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001
n = 115 n = 116 n = 115 n = 116
AUC: Area-under-the-curve; CK: creatine kinase, ESCALAT: Efegatran and Streptokinase to Canalize Arteries Like Accelerated Tissue plasminogen activator, LVEF: left ventricular ejection fraction, PRIME: Prompt Reperfusion In Myocardial infarction Evolution, SPECT: single-photon emission computed tomography
Figure 3 Natural log of peak CK-MB and percent infarcted myocardium (SPECT). AUC: Area-under-the-curve, CK: Creatine kinase, LN: log normal, SPECT: Single-photon emission computed tomography.
Correlation of enzymes and LVEF
In PRIME, repeat angiography with LVEF assessment was repeated after a median (25th, 75th) of 5 days [4,6] from thrombolysis. A statistically significant negative correlation existed between LVEF measured at repeat angiography and the peak levels of CK (r = -0.30, P = 0.035). Borderline significant correlations were observed between LVEF and CK-AUC (r = -0.28, P = 0.051). The negative correlation between LVEF and peak CK-MB (r = -0.26, P = 0.06) as well as between LVEF and CK-MB AUC (r = -0.22, P = 0.11) did not achieve statistical significance (Table 2).
Association between enzymes and outcomes
The total number of events in each trial is shown in Table 3. Figure 4 displays a graph of death/CHF and peak CK/CK-AUC quintiles. Figure 5 is a similar graph of death/CHF and peak CK-MB /CK-MB AUC quintiles. The Wilcoxon rank sum test resulted in a statistically significant association between death/CHF and peak CK (P = 0.007) and AUC (P = 0.001) measurements; there was also an association between death/CHF and peak CK-MB (P = 0.021) and AUC (P = 0.005) measurements.
Table 3 ESCALAT and PRIME trials: clinical outcomes
ESCALAT n = 136 PRIME n = 165
Death 2 (1.5) 5 (3.0)
New CHF Not collected 21 (12.7)
Death/new CHF 2 (1.5) 22 (13.3)
Values are n (%)
CHF: Congestive heart failure, ESCALAT: Efegatran and Streptokinase to Canalize Arteries Like Accelerated Tissue plasminogen activator, PRIME: Prompt Reperfusion In Myocardial infarction Evolution
Figure 4 Incidence of new CHF and death by CK-AUC and peak CK measurements, according to quintiles. AUC: Area-under-the-curve, CHF: congestive heart failure, CK: Creatine kinase.
Figure 5 Incidence of new CHF and death by CK-MB AUC and peak CK-MB measurements, according to quintiles. AUC: Area-under-the-curve, CHF: congestive heart failure, CK: Creatine kinase.
Discussion
This study showed a significant correlation between infarct size as assessed by SPECT and infarct size as determined by CK and CK-MB release using AUC or peak values. Weaker correlations were observed between infarct size measured by cardiac biomarker assessments and LVEF determined by angiography. Patients with larger release of CK or CK-MB had an increased risk of death and/or CHF.
These findings are consistent with earlier studies [12,24-33] that have also examined the correlation between CK and CK-MB and SPECT-determined infarct size. However, the majority of these analyses were done in the era prior to reperfusion therapy with fibrinolytic therapy or percutaneous coronary intervention. Therefore, the current analysis is important for trials being planned in the STE AMI population with contemporary therapies.
Attempts have been made to correlate CK- and CK-MB-based infarct measurements with LV function, with conflicting results. EF has been variably correlated to CK-MB AUC [17,34], peak CK-MB [17,35], CK AUC [15,36], and peak CK [36] in several angiographic studies, as well as in 1 radionucleotide study involving patient numbers similar to those in our study. Although the data confirm a significant association between LVEF determined by angiography and peak CK, the Spearman coefficient revealed only a weak relationship. This may have been in part due to small sample sizes.
The relatively poor correlation noted here between angiographic and enzymatic methods of infarct sizing might be due to post-infarct stunning, leading to an artifactually low LVEF; earlier clinically or electrocardiographically silent MIs contributing to LV dysfunction attributed to the index event; site of infarct (inferior vs. anterior); and the degree of right ventricular involvement. There also appears to be wide variability in the published Spearman coefficients on enzyme/LVEF correlations, with the current data falling at the lower end of the previously published ranges. These discrepant results might be explained by a number of factors, including variations in enzyme curve measurements; inter- and intra-study variability of LVEF measurements; timing of catheterization relative to MI; differences in treatment allotments; the year of study, which has implications for the availability of reperfusion and improvements in coronary care; and publication bias, with nonsignificant or weak association preferentially not being reported in the literature.
The ability of cardiac enzymes to predict clinical outcomes would enhance confidence in the use of this assessment in phase II clinical trials but would not exclude the need to conduct definitive trials designed to assess clinical outcomes directly. An early study of patients presenting with acute MI prior to the era of widespread revascularization showed a correlation between peak CK and CK-AUC and arrhythmias, pulmonary edema, and higher pulmonary artery wedge pressure [37]. Subsequently, in the Global Utilization of Streptokinase and Tissue plasminogen activator for Occluded coronary arteries (GUSTO-I) enzyme substudy [38], a 12% decrease in infarct size (as measured by α-hydroxybutyrate dehydrogenase), with t-PA (compared with streptokinase) corresponded to the 14% relative mortality reduction in the trial overall [39]. More recently, analysis from the Thrombolysis and Angioplasty in Myocardial Infarction (TAMI-7) trial of accelerated alteplase in patients presenting with ST-segment elevations showed a clear correlation between CK-MB peak and AUC values and development of CHF and a composite of CHF and death [17]. Similarly, data from the Platelet glycoprotein IIb/IIIa in Unstable angina: Receptor Suppression Using Integrilin Therapy (PURSUIT) trial showed a graded risk of death at 30 days and 6 months based on peak CK and CK-MB in NSTEMI [19]. In keeping with prior studies, we were able to demonstrate a graded risk of adverse cardiovascular outcomes (defined as new CHF and death) with a statistically significant increase of events in patients in higher quintiles of peak and AUC measurements for both CK and CK-MB.
The strengths of the present study are that the data were collected in the context of 2 randomized, controlled trials with strict adherence to serial biomarker assessments, analyses by a central core laboratory, and sophisticated curve-fitting techniques. Most of the noted previous studies used the modified formula proposed by Roberts et al [2] in 1975, to calculate the cumulative enzyme release. For this study, in contrast, a log normal function was used to correlate AUC values and to curve-fit in order to find peak enzyme levels; more recent data show that the log normal function best approximates the true enzyme release curve and peak values [23]. This method is easy to use and requires less sampling than other methods of curve fitting. The log normal function has obvious advantages over older methods and a clear role in data collection for prospective studies of infarction.
Limitations
Several weaknesses should be acknowledged. Sample sizes from the 2 trials were relatively small. The ventriculographic and scintigraphic data were incomplete, which may have implications if the missing data were significantly different from that reported here. Also, to more accurately assess the damage done by the index infarction, this analysis included only patients presenting with their first MI. This factor may affect the generalizability of the results. Although plot-fitting appears to be the most accurate method of determining enzyme peaks and integrating areas given current routine practices in measuring cardiac markers, it is possible that the sampling frequency used misrepresented the true nature of the enzyme curves. Finally, cardiac biomarkers, though useful, still represent surrogate measurements of clinical outcomes and need to be used with caution in the evaluation of new therapies, because a new treatment that has had no effect on a surrogate measurement (such as CK or CK-MB) might still have an important treatment effect.
Given the small sample size, we were not able to rigorously assess the potential impact of infarct location on the relationships observed. The contribution of right ventricular infarct to CK-MB release and the correlation of this release to assessment of left ventricular function could confound these findings.
Implications
Given their affordability and routine availability, cardiac biomarkers remain an attractive tool for measuring myocardial damage and determining prognosis. Cardiac enzyme markers continue to be used as endpoints in clinical trials, such as the Complement And ReDuction of INfarct size after Angioplasty or Lytics (CARDINAL) program [40,41] and the Limitation of Myocardial Infarction Following Thrombolysis in Acute Myocardial Infarction (LIMIT AMI) [42] trials. These results support the use of biomarker-based infarct sizing in the evaluation of therapeutics, treatment effects, and outcomes.
An attractive alternative in early drug development is to continue to combine a variety of measures of efficacy, such as cardiac enzymes, ECG analysis, SPECT, and angiography, to develop a "biomarker array." If multiple validated and significant prespecified outcomes are analyzed in the context of phase II clinical research, fewer patients would potentially need to be enrolled to discern differences in treatment effects between trial arms. This analysis substantiates the concept that routine cardiac markers can predict infarct size and clinical outcomes and may therefore act as a valid endpoint in phase II trials of acute MI therapy.
Conclusion
Serial cardiac biomarker assessment and curve-fitting techniques can be used to determine infarct size with statistically significant correlation to SPECT imaging. Weaker correlations were observed with LVEF by angiography. Death and CHF outcomes were associated with larger infarct size as determined by biomarkers. These data support the use of biomarker-determined infarct size as a potential endpoint in phase II clinical trials.
Abbreviations
AUC, area-under-the-curve
CHF, congestive heart failure
CK, creatine kinase
LVEF, left ventricular function
MI, myocardial infarction
NSTEMI non-ST-segment myocardial infarction
SPECT, single-photon emission computed tomography
STEMI, ST-segment myocardial infarction
t-PA, tissue plasminogen activator
Competing interests
The authors have no competing interests to declare.
The PRIME trial was funded by grants from Centocor, Inc., Malvern, Pennsylvania, and Eli Lilly and Company, Indianapolis, Indiana. The ESCALAT trial was sponsored by Lilly Research Laboratories, Indianapolis, Indiana; Eli Lilly Canada, Inc., Scarborough, Ontario, Canada; and LillyResearch Center Ltd., Surrey, United Kingdom.
Authors' contributions
ATT served as a primary author. KWM served as a primary author and provided critical review of manuscript. DG provided statistical analyses. WDW provided critical review of manuscript. RHC provided critical review of manuscript. NRE provided critical review of manuscript. EMO provided critical review of manuscript. All authors have given final approval of the version to be published.
Acknowledgements
The PRIME trial was funded by grants from Centocor, Inc., Malvern, Pennsylvania, USA, and Eli Lilly and Company, Indianapolis, Indiana, USA. The ESCALAT trial was sponsored by Lilly Research Laboratories, Indianapolis, Indiana, USA; Eli Lilly Canada, Inc., Scarborough, Ontario, Canada; and Lilly Research Center Ltd., Surrey, United Kingdom.
Vivian McGee provided editing services on behalf of the Duke Clinical Research Institute.
==== Refs
Shell WE Lavelle JF Covell JW Sobel BE Early estimation of myocardial damage in conscious dogs and patients with evolving acute myocardial infarction J Clin Invest 1973 52 2579 2590 4729051
Roberts R Henry PD Sobel BE An improved basis for enzymatic estimation of infarct size Circulation 1975 52 743 754 1236776
Norris RM Whitlock RM Barratt-Boyes C Small CW Clinical measurement of myocardial infarct size. Modification of a method for the estimation of total creatine phosphokinase release after myocardial infarction Circulation 1975 51 614 620 1116251
Roe CR Cobb FR Starmer CF The relationship between enzymatic and histologic estimates of the extent of myocardial infarction in conscious dogs with permanent coronary occlusion Circulation 1977 55 438 449 837480
Poliner LR Buja LM Parkey RW Stokely EM Stone MJ Harris R Saffer SW Templeton GH Bonte FJ Willerson JT Comparison of different noninvasive methods of infarct sizing during experimental myocardial infarction J Nucl Med 1977 18 517 523 870641
Jarmakani JM Limbird L Graham TC Marks RA Effect of reperfusion on myocardial infarct, and the accuracy of estimating infarct size from serum creatine phosphokinase in the dog Cardiovasc Res 1976 10 245 253 938992
Roe CR Starmer CF A sensitivity analysis of enzymatic estimation of infarct size Circulation 1975 52 1 5 1132111
Roe CR Validity of estimating myocardial infarct size from serial measurements of enzyme activity in the serum Clin Chem 1977 23 1807 1812 902405
Shell WE Kjekshus JK Sobel BE Quantitative assessment of the extent of myocardial infarction in the conscious dog by means of analysis of serial changes in serum creatine phosphokinase activity J Clin Invest 1971 50 2614 2625 5129313
Voss EM Sharkey SW Gernert AE Murakami MM Johnston RB Hsieh CC Apple FS Human and canine cardiac troponin T and creatine kinase-MB distribution in normal and diseased myocardium. Infarct sizing using serum profiles Arch Pathol Lab Med 1995 119 799 806 7668937
Grande P Christiansen C Alstrup K Comparison of ASAT, CK, CK-MB, and LD for the estimation of acute myocardial infarct size in man Clin Chim Acta 1983 128 329 335 6851141 10.1016/0009-8981(83)90332-7
Grande P Hansen BF Christiansen C Naestoft J Estimation of acute myocardial infarct size in man by serum CK-MB measurements Circulation 1982 65 756 764 7060254
Grande P Kiilerich S Relationship between serum CK-MB-estimated acute myocardial infarct size and clinical complications Acta Med Scand 1984 215 355 362 6731045
Grande P Pedersen A Myocardial infarct size: correlation with cardiac arrhythmias and sudden death Eur Heart J 1984 5 622 627 6479189
Isobe M Nagai R Ueda S Tsuchimochi H Nakaoka H Takaku F Yamaguchi T Machii K Nobuyoshi M Yazaki Y Quantitative relationship between left ventricular function and serum cardiac myosin light chain I levels after coronary reperfusion in patients with acute myocardial infarction Circulation 1987 76 1251 1261 3677350
Grande P Christiansen C Pedersen A Influence of acute myocardial infarct size on acute and one-year mortality Eur Heart J 1983 4 20 25 6832175
Christenson RH Vollmer RT Ohman EM Peck S Thompson TD Duh SH Ellis SG Newby LK Topol EJ Califf RM Relation of temporal creatine kinase-MB release and outcome after thrombolytic therapy for acute myocardial infarction. TAMI Study Group Am J Cardiol 2000 85 543 547 11078264 10.1016/S0002-9149(99)00808-5
Hindman N Grande P Harrell FE JrAnderson C Harrison D Ideker RE Selvester RH Wagner GS Relation between electrocardiographic and enzymatic methods of estimating acute myocardial infarct size Am J Cardiol 1986 58 31 5 3728328 10.1016/0002-9149(86)90236-5
Alexander JH Sparapani RA Mahaffey KW Deckers JW Newby LK Ohman EM Corbalan R Chierchia SL Boland JB Simoons ML Califf RM Topol EJ Harrington RA Association between minor elevations of creatine kinase-MB level and mortality in patients with acute coronary syndromes without ST-segment elevation. PURSUIT Steering Committee JAMA 2000 283 347 353 10647797 10.1001/jama.283.3.347
Fung AY Lorch G Cambier PA Hansen D Titus BG Martin JS Lee JJ Every NR Hallstrom AP Stock-Novack D Scherer J Weaver WD Efegatran sulfate as an adjunct to streptokinase versus heparin as an adjunct to tissue plasminogen activator in patients with acute myocardial infarction. ESCALAT Investigators Am Heart J 1999 138 696 704 10502216
PRIME Investigators Multicenter, dose-ranging study of efegatran sulfate versus heparin with thrombolysis for acute myocardial infarction. The Promotion of Reperfusion in Myocardial Infarction Evolution (PRIME) trial Am Heart J 2002 143 95 105 11773918 10.1067/mhj.2002.119896
Hurrell DG Milavetz J Hodge DO Gibbons RJ Infarct size determination by technetium 99 m sestamibi single-photon emission computed tomography predicts survival in patients with chronic coronary artery disease Am Heart J 2000 140 61 66 10874264 10.1067/mhj.2000.105104
Vollmer RT Christenson RH Reimer K Ohman EM Temporal creatine kinase curves in acute myocardial infarction. Implications of a good empiric fit with the log-normal function Am J Clin Pathol 1993 100 293 298 8379538
Licka M Zimmermann R Zehelein J Dengler TJ Katus HA Kubler W Troponin T concentrations 72 hours after myocardial infarction as a serological estimate of infarct size Heart 2002 87 520 524 12010931 10.1136/heart.87.6.520
Wagner I Mair J Fridrich L Artner-Dworzak E Lechleitner P Morass B Dienstl F Puschendorf B Cardiac troponin T release in acute myocardial infarction is associated with scintigraphic estimates of myocardial scar Coron Artery Dis 1993 4 537 544 8261232
Fletcher JW Mueller HS Rao PS Sequential thallium-201 myocardial scintigraphy after acute infarction in man: comparison to creatine kinase-MB release pattern Radiology 1980 136 191 195 7384499
Tamaki S Nakajima H Murakami T Yui Y Kambara H Kadota K Yoshida A Kawai C Tamaki N Mukai T Ishii Y Torizuka K Estimation of infarct size by myocardial emission computed tomography with thallium-201 and its relation to creatine kinase-MB release after myocardial infarction in man Circulation 1982 66 994 1001 6982116
Tamaki S Murakami T Kadota K Kambara H Yui Y Nakajima H Suzuki Y Nohara R Takatsu Y Kawai C Effects of coronary artery reperfusion on relation between creatine kinase-MB release and infarct size estimated by myocardial emission tomography with thallium-201 in man J Am Coll Cardiol 1983 2 1031 1038 6415143
Mair J Wagner I Morass B Fridrich L Lechleitner P Dienstl F Calzolari C Larue C Puschendorf B Cardiac troponin I release correlates with myocardial infarction size Eur J Clin Chem Clin Biochem 1995 33 869 872 8620066
Jansen DE Corbett JR Wolfe CL Lewis SE Gabliani G Filipchuk N Redish G Parkey RW Buja LM Jaffe AS Quantification of myocardial infarction: a comparison of single photon-emission computed tomography with pyrophosphate to serial plasma MB-creatine kinase measurements Circulation 1985 72 327 333 2988821
Angeja BG Gunda M Murphy SA Sobel BE Rundle AC Syed M Asfour A Borzak S Gourlay SG Barron HV Gibbons RJ Gibson CM TIMI myocardial perfusion grade and ST segment resolution: association with infarct size as assessed by single photon emission computed tomography imaging Circulation 2002 105 282 285 11804979 10.1161/hc0302.103588
Sharpe DN Botvinick EH Shames DM Norman A Chatterjee K Parmley WW The clinical estimation of acute myocardial infarct size with 99mTechnetium pyrophosphate scintigraphy Circulation 1978 57 307 313 618619
Mahmarian JJ Pratt CM Borges-Neto S Cashion WR Roberts R Verani MS Quantification of infarct size by T1 single-photon emission computed tomography during acute myocardial infarction in humans Circulation 1988 78 831 839 3262453
Rogers WJ McDaniel HG Smith LR Mantle JA Russel RO JrRackley CE Correlation of angiographic estimates of myocardial infarct size and accumulated release of creatine kinase MB isoenzyme in man Circulation 1977 56 199 205 872311
Apple FS Sharkey SW Falahati A Murakami M Mitha N Christensen D Assessment of left ventricular function using serum cardiac troponin I measurements following myocardial infarction Clin Chim Acta 1998 272 59 67 9581857 10.1016/S0009-8981(97)00252-0
Blanke H von Hardenberg D Cohen M Kaiser H Karsch KR Holt J Smith H JrRentrop P Patterns of creatine kinase release during acute myocardial infarction after nonsurgical reperfusion: comparison with conventional treatment and correlation with infarct size J Am Coll Cardiol 1984 3 675 680 6693639
Ryan W Karliner JS Gilpin EA Covell JW DeLuca M Ross J Jr The creatine kinase curve area and peak creatine kinase after acute myocardial infarction: usefulness and limitations Am Heart J 1981 101 162 168 7468417 10.1016/0002-8703(81)90660-8
Baardman T Hermens WT Lenderink T Molhoek GP Grollier G Pfisterer M Simoons ML Differential effects of tissue plasminogen activator and streptokinase on infarct size and on rate of enzyme release: influence of early infarct related artery patency. The GUSTO Enzyme Substudy Eur Heart J 1996 17 237 246 8732377
The GUSTO Investigators An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction N Engl J Med 1993 329 673 682 8204123 10.1056/NEJM199309023291001
Mahaffey KW Granger CB Nicolau JC Ruzyllo W Weaver WD Theroux P Hochman JS Filloon TG Mojcik CF Todaro TG Armstrong PW COMPLY Investigators Effect of pexelizumab, an anti-C5 complement antibody, as adjunctive therapy to fibrinolysis in acute myocardial infarction: the COMPlement inhibition in myocardial infarction treated with thromboLYtics (COMPLY) trial Circulation 2003 108 1176 1183 12925455 10.1161/01.CIR.0000087404.53661.F8
Granger CB Mahaffey KW Weaver WD Theroux P Hochman JS Filloon TG Rollins S Todaro TG Nicolau JC Ruzyllo W Armstrong PW COMMA Investigators Pexelizumab, an anti-C5 complement antibody, as adjunctive therapy to primary percutaneous coronary intervention in acute myocardial infarction: the COMplement inhibition in Myocardial infarction treated with Angioplasty (COMMA) trial Circulation 2003 108 1184 1190 12925454 10.1161/01.CIR.0000087447.12918.85
Baran KW Nguyen M McKendall GR Lambrew CT Dykstra G Palmeri ST Gibbons RJ Borzak S Sobel BE Gourlay SG Rundle AC Gibson CM Barron HV LIMIT AMI Study Group Double-blind, randomized trial of an anti-CD18 antibody in conjunction with recombinant tissue plasminogen activator for acute myocardial infarction: limitation of myocardial infarction following thrombolysis in acute myocardial infarction (LIMIT AMI) study Circulation 2001 104 2778 2783 11733394
|
16115321
|
PMC1236947
|
CC BY
|
2021-01-04 16:47:32
|
no
|
Curr Control Trials Cardiovasc Med. 2005 Aug 23; 6(1):12
|
utf-8
|
Curr Control Trials Cardiovasc Med
| 2,005 |
10.1186/1468-6708-6-12
|
oa_comm
|
==== Front
Environ HealthEnvironmental Health1476-069XBioMed Central London 1476-069X-4-181612902610.1186/1476-069X-4-18ResearchHazardous waste sites and stroke in New York State Shcherbatykh Ivan [email protected] Xiaoyu [email protected] Lawrence [email protected] David O [email protected] Department of Environmental Health and Toxicology, School of Public Health, University at Albany, SUNY, One University Place, A217, Rensselaer, NY 12144, USA2 Department of Biometry and Statistics, School of Public Health, University at Albany, SUNY, One University Place, A217, Rensselaer, NY 12144, USA3 Institute for Health and the Environment, University at Albany, SUNY, One University Place, A217, Rensselaer, NY 12144, USA4 McMaster University, Centre for Evaluation of Medicines, 105 Main St. E., P1 Level, Hamilton, Ontario L8N 1G6, Canada2005 29 8 2005 4 18 18 28 4 2005 29 8 2005 Copyright © 2005 Shcherbatykh et al; licensee BioMed Central Ltd.2005Shcherbatykh 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 -
Environmental exposure to persistent organic pollutants (POPs) may lead to elevation of serum lipids, increasing risk of atherosclerosis with thromboembolism, a recognized cause of stroke. We tested the hypothesis that exposure to contaminants from residence near hazardous waste sites in New York State influences the occurrence of stroke.
Methods -
The rates of stroke hospital discharges were compared among residents of zip codes containing hazardous waste sites with POPs, other pollutants or without any waste sites using information for 1993–2000 from the New York Statewide Planning and Research Cooperative System (SPARCS) database, containing the records of all discharge diagnoses for patients admitted to state-regulated hospitals.
Results -
After adjustment for age and race, the hospitalization rate for stroke in zip codes with POPs-contaminated sites was 15% higher than in zip codes without any documented hazardous waste sites (RR 1.15, 95% CI, 1.05, 1.26). For ischemic stroke only, the RR was 1.17 (95% CI 1.04, 1.31). Residents of zip codes containing other waste sites showed a RR of 1.13 (95% CI, 1.02, 1.24) as compared to zip codes without an identified waste site.
Conclusion -
These results suggest that living near a source of POPs contamination constitutes a risk of exposure and an increased risk of acquiring cerebrovascular disease. However further research with better control of individual risk factors and direct measurement of exposure is necessary for providing additional support for this hypothesis.
==== Body
Background
Cerebrovascular disease is a major public health problem [1]. In addition to well-documented modifiable risk factors of stroke, there is evidence for a link between a broad category of environmental factors and stroke, such as air pollution [2], environmental tobacco smoke [3], metals [4], pesticides [5], and other anthropogenic factors, including persistent organic pollutants (POPs).
POPs are chlorinated organic compounds [polychlorinated biphenyls (PCBs), dioxins and chlorinated pesticides] that are resistant to degradation and able to bio-accumulate in fatty tissues of living organisms. These compounds are semivolatile, and present in the atmosphere as vapors or adsorbed on suspended particles [6]. Multiple adverse health effects have been associated with the exposure to POPs [7,8] including cardiovascular pathologies, such as hypertension, ischemic heart disease, and atherosclerosis [9-11]. A major route of exposure to POPs is dietary, especially through the consumption of fish from contaminated waters. There is, however, growing evidence that suggests inhalation is a significant route of exposure to POPs and, especially, to PCBs [12,13]. Waste combustion and volatilization from wet soils or sewage sludge-amended land considerably contribute to elevation of PCB levels in the environment [14] and increase a possibility of human exposure through inhalation [15,16].
There is an increased incidence of some chronic diseases among individuals living near hazardous waste sites [17]. However, to our knowledge there have been no studies that explored an association between xenobiotic exposure from hazardous waste sites and cerebrovascular disease. We have performed a study using aggregate data to investigate the relationship between environmental exposure to POPs/PCBs and other wastes and hospital discharge rate of stroke among the New York State residents.
Methods
SPARCS was used to obtain data on hospital discharge diagnosis of cerebrovascular disease among the New York State residents. SPARCS contains records of discharge diagnoses for all persons admitted as inpatients in all public and private New York hospitals, excluding federal-regulated facilities and mental health facilities. Non-hospitalized cases are not collected by the SPARCS registry. The database available to us contained the primary and up to 14 secondary discharge diagnoses in the format of International Classification of Disease, Ninth Revision (ICD-9), in addition to the zip code of residence, sex, age, and race/ethnicity of each patient. We have used the SPARCS data from 1993 to 2000.
Data on hazardous waste sites were obtained from the New York State Department of Environmental Conservation (NYSDEC). The major contaminants present and the zip code(s) for each site were extracted from the NYSDEC database. From a total of 818 State and Federal hazardous waste sites in New York State (excluding New York City), 396 sites contained POPs. These hazardous waste sites ("POPs") were located in 192 zip codes. Two hundred thirteen zip codes contained hazardous waste sites where the listed contaminants of concern did not include any POP, and these were categorized as "other waste". All of the other 994 zip codes, which contained no identified hazardous waste sites, were classified as "clean", although we recognize that these zip codes may contain wastes that have not been characterized. One subset of the "POPs sites" was examined separately: 78 zip codes along the PCB-contaminated portion of the Hudson River from Hudson Falls to New York City (30% of all people living in POPs-contaminated zip codes). Using the information from the Behavioral Risk Factor Surveillance System (BRFSS) we were able to compare behavior of the population along the Hudson River (available only at the level of counties, not zip codes) to the rest of New York State.
Demographic data on New York State residents (age, gender, race, and median household income for each zip code in Upstate New York) was obtained from Claritas Inc., an information resource company that provides information derived from the U.S. Census, with the zip codes used being the same as those used by the U.S. Postal Service.
The data from SPARCS, Claritas, and NYSDEC were merged on the basis of a zip code of residence to determine the rate of stroke hospital discharges among the individuals residing in three categories of zip codes ("POPs", "other waste", and "clean") for the years 1993–2000. Exposure was defined as a patient's residence in a zip code that contained or abutted at least one hazardous waste site. We used primary and all secondary ICD-9 codes 430 to 436 for cerebrovascular disease (with the fourth and fifth digits). Ischemic stroke was defined as codes 433.x1, 434.x1, and 436, while hemorrhagic cerebrovascular disease was identified as codes 430, 431, and 432.
We excluded all zip codes that were not constant from 1993 to 2000 and post office box zip codes (914 in total). Since New York City maintains its own hospitalization dataset and has unique sociodemographic characteristics, it was also excluded. After all the exclusions were made, 1399 zip codes remained in the study.
Finally, the analysis was restricted to White and African-American races because the numbers of Asians and Native Americans were small. We restricted analysis to patients between 25 to 64 years old in order to evaluate stroke frequency at an age at which stroke is relatively rare, expecting that that would provide a better indication of an elevation in risk should it exist.
Statistical analysis
The stroke hospital discharge rate per 100,000 was calculated as the number of people discharged with cerebrovascular disease divided by the estimated total population. We used a Negative Binomial regression model, with the GENMOD procedure from SAS software. The Negative Binomial model was log linear: Log (Expected Hospital Discharge Rate of Stroke) = log (total person-time) + β0 + β1*AGE5 + β2*AGE4 + β3*AGE3 + β4*GENDER + β5*RACE + β6*POPs sites + β7*Other waste, where "POPs" and "other waste" represented the exposure; age, gender, and race represented other dependent variables with a value of zero or one. Before formulating the final regression model, we assessed confounding by demographic variables (age, gender, and race). All statistical analyses were conducted using the SAS statistical software package, version 8.2 (SAS Institute, Inc., Cary, North Carolina).
The initial Negative Binomial regression analysis included all four quartiles of the median household income, estimated on a zip code level. However, the analysis of zip codes with the lowest and highest median incomes (the first and the fourth quartiles) showed the greatest population variability. Therefore, we restricted the Negative Binomial regression model to the middle-income zip codes (second and third quartiles), with the median household income ranging from $30,388.0 to $48,213.5.
Results
Figure 1 shows the crude rates of hospital discharge for ischemic and hemorrhagic stroke in the full dataset of 42,420,284 person-years over 1993–2000. For ischemic stroke there was a significantly greater number of discharges in residents of "POPs" sites as compared to either "clean" or "other waste" sites (P < 0.0001). For hemorrhagic stroke the elevation was significant in "POPs" sites as compared to "clean" sites (P = 0.003).
Figure 1 Crude hospital discharge rates for ischemic and hemorrhagic stroke in upstate New York, 1993–2000, for residents living in clean, POPs, and other waste site zip codes.
From 1993 to 2000 there were 28,216 stroke discharges in the three study zip code classes. Table 1 shows the population characteristics of the second and third income quartiles that were used for the further analysis. Such variables as age, gender, race and median income did not differ significantly among "POPs", "other waste" and "clean" zip codes. The age, gender and race distribution of individuals with stroke discharges and residing in the three groups of zip code are presented in Table 2.
Table 1 Population Characteristics*
Person-years (%)
"POPs" zip codes "Other waste" zip codes "Clean" zip codes
Total† (16,880,516) 5,815,148 (34.5) 4,340,180 (25.7) 6,725,188 (39.8)
Average Income‡ $37,971.09 $36,547.13 $36,231.88
Age groups
25–34 1,703,276 (29.3) 1,261,376 (29.1) 1,889,316 (28.1)
35–44 1,774,672 (30.5) 1,323,224 (30.5) 2,070,400 (30.8)
45–54 1,335,396 (23.0) 1,022,544 (23.6) 1,625,412 (24.2)
55–64 1,001,804 (17.2) 733,036 (16.9) 1,140,060 (16.9)
Gender
Male 2,858,752 (49.2) 2,142,968 (49.4) 3,371,492 (50.1)
Female 2,956,396 (50.8) 2,197,212 (50.6) 3,353,696 (49.9)
Race
White 5,361,748 (92.2) 3,934,472 (90.7) 6,425,104 (95.5)
African American 453,400 (7.8) 405,708 (9.3) 300,084 (4.5)
* This and all subsequent tables represent data for 1993–2000
† Total population is the total person-years from 1993 through 2000, which is the sum of the population in each category of zip codes each year for eight years.
‡ Average income was calculated as the sum of the median household income for each zip code divided by the number of zip codes in each category.
Table 2 Age Distribution for Stroke Discharges
Stroke Discharges N (%)
"POPs" zip codes "Other Waste" zip codes "Clean" zip codes
Total Discharges* 10,220 7,506 10,490
Age groups
25–34 322 (3.2†) 238 (3.2) 287 (2.7)
35–45 986 (9.6) 683 (9.1) 943 (9.0)
45–54 2,768 (27.1) 2,069 (27.5) 2,874 (27.4)
55–64 6,144 (60.1) 4,516 (60.2) 6,386 (60.9)
* The total stroke discharges in all three categories of zip codes was 28,216.
† The percentage of stroke discharges in the "POPs" sites in the age group 25–34 years.
The results of negative binomial regression are presented in Table 3. The rate ratio (RR) for stroke discharge in zip codes with POPs-contaminated hazardous waste sites was 1.15 (95% CI 1.05, 1.26) as compared to zip codes without any contamination. We also observed a RR of 1.13 (95% CI 1.02, 1.24) in "other waste" sites as compared to "clean" sites. As expected, age, race, and gender were significantly associated with the stroke discharges. The goodness-of-fit criteria demonstrated an adequate fitting of the model.
Table 3 Results of Negative Binomial Regression for All Types of Stroke
Rate Ratio 95% Confidence Intervals for Rate Ratio
Exposure
"POPs" zip codes 1.15 1.05 1.26
"Other waste" zip codes 1.13 1.02 1.24
"Clean" zip codes* 1.00 1.00 1.00
Age Groups
55–64 33.28 29.50 37.54
45–54 11.08 9.81 12.52
35–44 3.04 2.68 3.46
25–34* 1.00 1.00 1.00
Race
African-American 2.06 1.90 2.24
White* 1.00 1.00 1.00
Gender
Male 1.13 1.04 1.23
Female* 1.00 1.00 1.00
* Reference group
Table 4A and 4B show results of negative binomial regression for ischemic and hemorrhagic stroke. Rates for ischemic stroke were elevated in zip codes with sites containing POPs (RR 1.17, 95% CI 1.04, 1.31) and other waste (RR 1.14, 95% CI 1.01, 1.28), as compared to "clean" zip codes. For hemorrhagic stroke the RRs were not significantly different. Age, race, and gender were significantly associated with the hospital discharges of both types of stroke (data not shown).
Table 4 Negative binomial regression results for stroke, and comparison of Hudson River to "clean" zip codes.
Exposure Rate Ratio 95% Confidence Interval
A. Ischemic Stroke
"POPs" zip codes 1.17 1.04 1.39
"Other waste" zip codes 1.14 1.01 1.28
"Clean" zip codes* 1.00 1.00 1.00
B. Hemorrhagic Stroke
"POPs" zip codes 1.10 0.99 1.22
"Other waste" zip codes 1.04 0.93 1.16
"Clean" zip codes* 1.00 1.00 1.00
C. Subset of Hudson River zip codes†
Hudson River zip codes 1.20 1.10 1.32
"Clean" zip codes* 1.00 1.00 1.00
* Zip codes with no identified hazardous waste sites (reference)
† Zip codes along the PCB-contaminated portion of the Hudson River
Table 4C shows the negative binomial regression model results for stroke for a subset of POPs zip codes, those along the Hudson River. As compared to "clean" zip codes, the RR was 1.20 (95% CI 1.10, 1.32). This 20% elevation of stroke diagnosis in Hudson River zip codes indicates that the relationship between residential exposure to POPs and the rate of stroke discharges is similar for the people living along the Hudson River and general population of Upstate New York.
Discussion
Our results suggest that living in zip codes that contain hazardous waste sites is associated with an increased rate of hospital discharges for stroke, especially ischemic stroke. The regression model, when limited to middle income, showed a 15% elevation of hospital discharge rates for stroke in zip codes with POPs waste sites, independent from patient's age, race, or gender even though increased age, being male and being African-American were all significant but independent risk factors. In POPs-contaminated zip codes along the Hudson River the increase of hospital discharge rates for stroke was even larger. However, as previously reported [18] individuals along the Hudson River on average have higher income, exercise more frequently, consume more fruits and vegetables, and there are more non-smokers and former smokers among them than among the people from the rest of Upstate New York. Therefore, these known risk factors alone cannot fully explain the observed geographical variation of stroke discharge rates. This provides some additional support for the hypothesis that exposure to POPs is a potential contributing factor. If the current data reflect true associations, it is unlikely that fish consumption, usually considered to be the primary route of exposure, is the only important one. Sport fishing and fish-eating habits are not defined by zip code of residency. Inhalation of POPs in the vapor phase or bound to particulates, or ingestion of POPs-contaminated particulates with foodstuff is the most likely exposure pathway.
The stroke discharge rates in zip codes that have a hazardous waste site, but not one with POPs, were also found to be elevated (12%) after the negative binomial regression analysis. However, the substantial heterogeneity in this group of pollutants (heavy metals, volatile organic substances, radiation, etc.) and existence of toxicological/biological interactions among them prevents us from drawing definite conclusions about their influence on stroke occurrence.
The results of this study are consistent with the findings of previous investigations that show elevations in various diseases in residents living near hazardous waste sites [2,13,17-19]. We [19] have reported an elevation is hospitalization for coronary artery disease and myocardial infarction using SPARCS data and the same characterization of zip codes as applied in this study. We found that residence in a POPs zip code in all of upstate New York resulted in a 15% elevated rate of hospitalization for coronary artery disease, and a 20% elevation in acute myocardial infarction. In the subset of POPs zip codes along the Hudson River the rates were elevated by 35.8% for coronary artery disease and by 39.1% for acute myocardial infarction.
Stroke has many pathologic factors similar to those in cardiovascular disorders. Ischemic stroke is a "brain attack" and its etiology is similar to that of myocardial infarction [20], while hemorrhagic stroke is often secondary to hypertension. Several studies have reported that exposure to POPs/PCBs is associated with elevated frequency of several chronic diseases, such as ischemic heart disease, diabetes, hypertension and chronic liver disease [9,21,22]. One mechanism linking exposure to POPs and cardiovascular diseases may be through increased serum lipids, a known risk factor for atherosclerosis. Serum lipids [23], plasma triglyceride levels [24], and total cholesterol [11] are elevated in dioxin and PCB-exposed populations. Monkeys exposed to dioxin and PCBs showed a 3- to 5-fold elevation in serum triglyceride concentrations [25], and similar effects have been reported in female rats after exposure to PCBs [26]. Dioxin and coplanar PCBs cause the production of reactive oxygen species which, in turn, cause damage to endothelial cells and promote the formation of foam cells and atherosclerotic plaques [27]. The combination of elevated serum lipids in the presence of damaged endothelial cells would contribute to the risk of development of cardiovascular pathology and cerebrovascular disease, especially ischemic stroke. Indeed, our results (Table 4A) show a stronger relation with ischemic than hemorrhagic stroke.
We have previously reported an elevation in hospital discharges for infectious respiratory disease in "POPs" zip code residents and compared to "clean" and "other waste" zip codes [18]. As in the present study, we found that the relationships were not reduced in zip codes along the Hudson River where income is higher and smoking rates lower than in the rest of New York State. Using a different data set, the New York State birth registry, we have also found an elevation in low birth weight in children born to residents of zip codes containing a PCB contaminated site [11]. In sum, together with our recent report on coronary heart disease and myocardial infarction [19], these results with four different health outcomes and two different datasets provide support for the hypothesis that residence near POPs sites increases risk of several types of disease.
There are clear limitations in determining cause and effect in partially ecologic study designs such as we have used. Given that the study was based on aggregative data, we do not have a direct measure of exposure (a zip code of residency is a crude surrogate for exposure assessment), and have no information on the duration of individual residence in each specific zip code. It is possible, and indeed likely, that some individuals residing in a POPs zip code were not exposed because of short length of residence, or residence still in the zip code but far from the hazardous waste site. The information on income was available only at a zip code level, which allowed marginal adjustment for socio-economic status. It does not adjust for the range of income in any one zip code, nor the possibility that within a zip code the income is less among those living closer to the waste sites. Therefore, it is possible that the observed associations were influenced by some underdetermined factors (e.g. access to health care) and might not be applicable to every socioeconomic subpopulation.
There are many other potential sources of confounding with the groups, and these are only partially controlled for by use of the BRFSS for the Hudson River population. BRFSS information is currently available only at a county level, not at the zip code level. While the BRFSS provides information on average behaviors within a county, it is still possible that those individuals who experience strokes differ from this average.
While the absence of personal identifiers restricted our ability to account for some potential confounders/effect modifiers, hospitalization data obtained in a mandatory manner for a number of years, because of the large numbers involved, has considerable potential for generating and testing hypotheses regarding the causes of disease [28]. In spite of these limitations our observations provide support for the general hypothesis that living near hazardous waste sites poses risk of exposure and of disease.
Conclusion
We found a statistically significant elevation of hospital discharge rates for stroke in zip codes with POPs-contaminated hazardous waste sites, and to a lesser degree with "other waste" sites, when compared to zip codes that do not have any identified waste sites. These observations suggest that living near a waste site contaminated with POPs is associated with the risk of inhalational and/or ingestional exposure and an increased risk of acquiring cerebrovascular disease. Further research involving control for individual risk factors and direct exposure assessment techniques is necessary for providing additional evidence for this hypothesis.
List of Abbreviations
BRFSS – Behavioral Risk Factor Surveillance System
ICD-9 – International Classification of Disease, Ninth Revision
NYSDEC – New York State Department of Environmental Conservation
PCBs – polychlorinated biphenyls
POPs – persistent organic pollutants
RR – rate ratio
SPARCS – New York Statewide Planning and Research Cooperative System
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
IS performed the data analysis as a requirement of his MPH program, performed the statistical analysis, and wrote the first draft of the paper. XH coordinated the use of the SPARCS and hazardous waste datasets. LL provided overall statistical direction for the study. DOC designed the study and supervised the data analysis. All authors read and approved the final manuscript.
Acknowledgements
Research for this article was supported in part by a grant from the Fogarty International Center 5D43TW00636 (to DOC), the Edmund S. Muskie/FREEDOM Support Act Graduate Fellowship Program, a program sponsored and managed by the Bureau of Educational and Cultural Affairs, U.S. Department of State (to IS), and by the Institute for Health and the Environment.
==== Refs
American Heart Association Heart Disease and Stroke Statistics – 2004 Update 2003 Dallas, TX: American Heart Association
Maheswaran R Elliott P Stroke mortality associated with living near main roads in England and Wales: a geographical study Stroke 2003 34 2776 2780 14615623 10.1161/01.STR.0000101750.77547.11
Bonita R Duncan J Truelsen T Jackson RT Beaglehole R Passive smoking as well as active smoking increases the risk of acute stroke Tob Control 1999 8 156 160 10478399
Salonen JT Seppanen K Nyyssonen K Korpela H Kauhanen J Kantola M Tuomilehto J Esterbauer H Tatzber F Salonen R Intake of mercury from fish, lipid peroxidation, and the risk of myocardial infarction and coronary, cardiovascular, and any death in eastern Finnish men Circulation 1995 91 645 655 7828289
Cantor KP Silberman W Mortality among aerial pesticide applicators and flight instructors: follow-up from 1965–1988 Am J Ind Med 1999 36 239 247 10398932 10.1002/(SICI)1097-0274(199908)36:2<239::AID-AJIM3>3.0.CO;2-V
Blais JM Froese KL Kimpe LE Muir DC Backus S Comba M Schindler DW Assessment and characterization of polychlorinated biphenyls near a hazardous waste incinerator: analysis of vegetation, snow, and sediments Environ Toxicol Chem 2003 22 126 133 12503755 10.1897/1551-5028(2003)022<0126:AACOPB>2.0.CO;2
Carpenter DO Polychlorinated biphenyls and human health Int J Occup Med Environ Health 1998 11 291 303 10028197
Weisglas-Kuperus N Neurodevelopmental, immunological and endocrinological indices of perinatal human exposure to PCBs and dioxins Chemosphere 1998 37 1845 1853 9828313 10.1016/S0045-6535(98)00250-1
Gustavsson P Hogstedt C A cohort study of Swedish capacitor manufacturing workers exposed to polychlorinated biphenyls (PCBs) Am J Ind Med 1997 32 234 249 9219652 10.1002/(SICI)1097-0274(199709)32:3<234::AID-AJIM8>3.0.CO;2-X
Pesatori AC Zocchetti C Guercilena S Consonni D Turrini D Bertazzi PA Dioxin exposure and non-malignant health effects: a mortality study Occup Environ Med 1998 55 126 131 9614398
Tokunaga S Kataoka K A longitudinal analysis on the association of serum lipids and lipoproteins concentrations with blood polychlorinated biphenyls level in chronic "Yusho" patients Fukuoka Igaku Zasshi 2003 94 110 117 12872711
Casey AC Berger DF Lombardo JP Hunt A Quimby F Aroclor 1242 inhalation and ingestion by Sprague-Dawley rats J Toxicol Environ Health A 1999 56 311 342 10094245 10.1080/009841099158033
Baibergenova A Kudyakov R Zdeb M Carpenter DO Low birth weight and residential proximity to PCB-contaminated waste sites Environ Health Perspect 2003 111 1352 1357 12896858
Sweetman AJ Jones KC Declining PCB concentrations in the U.K. atmosphere: Evidence and possible causes Environ Sci Technol 2000 34 863 869 10.1021/es9906296
Currado GM Harrad S A comparison of polychlorinated biphenyl concentrations in indoor and outdoor air and the potential significance of inhalation as a human exposure pathway Environ Sci Technol 1998 32 3043 3047 10.1021/es970735c
Voorhees DJ Cullen AC Altshul LM Exposure to polychlorinated biphenyls in residential indoor air and outdoor air near a superfund site Environ Sci Technol 1997 1 3612 3618 10.1021/es970371o
Vrijheid M Health effects of residence near hazardous waste landfill sites: a review of epidemiologic literature Environ Health Perspect 2000 108 101 112 10698726
Kudyakov R Baibergenova A Zdeb M Carpenter DO Respiratory disease in relation to patient residence near to hazardous waste sites Environ Toxicol Pharmacol 2004 18 249 257 10.1016/j.etap.2004.06.005
Seergev AV Carpenter DO Hospitalization rates for coronary heart disease in relation to residence near areas contaminated with persistent organic pollutants and other pollutants Environ Health Perspect 2005 113 756 761 15929900
Camarata PJ Heros RC Latchaw RE "Brain attack": the rationale for treating stroke as a medical emergency Neurosurgery 1994 34 144 157 8121551
Longnecker MP Klebanoff MA Brock JW Zhou H Polychlorinated biphenyl serum levels in pregnant subjects with diabetes Diabetes Care 2001 24 1099 1101 11375377
Stehr-Green PA Welty E Steele G Steinberg K Evaluation of potential health effects associated with serum polychlorinated biphenyl levels Environ Health Perspect 1986 70 255 259 3104024
Kreiss K Zack MM Kimbrough RD Needham LL Smrek AL Jones BT Association of blood pressure and polychlorinated biphenyl levels JAMA 1981 245 2505 2509 6785463 10.1001/jama.245.24.2505
Baker EL Landrigan PJ Glueck CJ Zack MM JrLiddle JA Burse VW Housworth WJ Needham LL Metabolic consequences of exposure to polychlorinated biphenyls (PCB) in sewage sludge Am J Epidemiol 1980 112 553 563 6775528
Rier SE Turner WE Martin DC Morris R Lucier GW Clark GC Serum levels of TCDD and dioxin-like chemicals in rhesus monkeys chronically exposed to dioxin: correlation of increased serum PCB levels with endometriosis Toxicol Sci 2001 59 147 159 11134554 10.1093/toxsci/59.1.147
Lind PM Orberg J Edlund UB Sjoblom L Lind L The dioxin-like pollutant PCB 126 (3,3',4,4',5-pentachlorobiphenyl) affects risk factors for cardiovascular disease in female rats Toxicol Lett 2004 150 293 299 15110081 10.1016/j.toxlet.2004.02.008
Hennig B Meerarani P Slim R Toborek M Daugherty A Silverstone AE Robertson LW Proinflammatory properties of coplanar PCBs: in vitro and in vivo evidence Toxicol Appl Pharmacol 2002 181 174 183 12079426 10.1006/taap.2002.9408
Tirschwell DL Longstreth WT Jr Validating administrative data in stroke research Stroke 2002 33 2465 2470 12364739 10.1161/01.STR.0000032240.28636.BD
|
16129026
|
PMC1236948
|
CC BY
|
2021-01-04 16:36:32
|
no
|
Environ Health. 2005 Aug 29; 4:18
|
utf-8
|
Environ Health
| 2,005 |
10.1186/1476-069X-4-18
|
oa_comm
|
==== Front
Harm Reduct JHarm Reduction Journal1477-7517BioMed Central London 1477-7517-2-121612238210.1186/1477-7517-2-12CommentaryThe rise of injecting drug use in east Africa: a case study from Kenya Beckerleg Susan [email protected] Maggie [email protected] Gillian Lewando [email protected] Public and Environmental Health Research Unit, London School of Hygiene & Tropical Medicine, London, UK2 Bristol Drugs Project, 11 Brunswick Square, Bristol, UK3 Social Sciences in Health, University of Warwick, UK2005 25 8 2005 2 12 12 11 7 2004 25 8 2005 Copyright © 2005 Beckerleg et al; licensee BioMed Central Ltd.2005Beckerleg 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.
Studies on injecting drug use in East Africa are reviewed. The existingstudies document the spread of heroin injection in Kenya and Tanzania, both countries where HIV rates are high. No data from Uganda on injecting drug use was found by the authors. A case study of the growth of heroin injection in a Kenyan coastal town is presented. The need for needle-exchange programmes and other prevention services is discussed.
==== Body
Background
Although bearing the brunt of the AIDS epidemic, Africa has long been considered largely free of injection drug use. Notwithstanding the assessments of the UN International Drug Control Programme [1] 1, international organisations have been slow to recognise either the spread of heroin use in Kenya or the existence of injection drug use. The largely unheeded spread of injection drug use in East Africa has wide implications for public health in the region. Injection drug users (IDU) are a 'high risk' or 'core group' for HIV infection. Many IDU share needles and syringes as well as having unprotected sex, and have been identified as a 'bridging population', speeding the spread of HIV to the general population [2,3] and [4].
Heroin injection now appears to be occurring in most large towns of Kenya and Tanzania. A study of 336 heroin users in Nairobi, Kenya found that 44.9% were, or had been, injectors [5]. Of 101 current injectors, 52.5% were HIV positive. This compares with a 13.5% prevalence rate among heroin users who had never injected. Hepatitis C prevalence also varied dramatically, from 61.4% among current injectors to 3.8% for those who had never injected. A similar study for Mombasa, Kenya's second city and main port, has been planned, but at the time of writing (November 2003) is yet to be carried out. However the UNODC and WHO are carrying out research in 2003 to establish links between HIV and drugs, with a focus on injecting, at the Kenya Coast.
Recent assessments in neighbouring Tanzania have found heroin injecting to be spreading throughout the country. Hence, a rapid situation assessment carried out in five Tanzanian towns [6] found heroin to be a major concern in Arusha, Dar es Salaam and Zanzibar, to be emerging as problem in Mwanza, but not in Mbeya. Injection drug use was reported in all the study sites where heroin was in use. Similarly, a study of 624 young multi-drug (alcohol, cannabis, tobacco, heroin, Valium, khat) users in Dar es Salaam found that 75% of the sample were using heroin, and that 114 (18.3%) of the sample reported injecting drugs [7]. As many of the substances used by the 624 people interviewed are not usually injected, the percentage of heroin users injecting in Dar es Salaam will be considerably higher than is indicated by these data which are not disaggregated by substance.
Much less is known about injecting drug use in Uganda. Indeed, the UNODC covering Eastern and Southern Africa reports that there have been no drug assessments carried out in Uganda.
Discussion
Case Study: The Kenya Coast
The case study draws together information collected by SB and MT from 1995 to 2003 as part of their work with the Omari Project, a Kenyan organisation pioneering the adaptation of international best practice treatment models for heroin users to the local setting of Coastal Kenya [8] and [9]. In addition, SB collected data on heroin users and injecting patterns between 2000 and 2001, as part of a larger ESRC funded study of women heroin users and their reproductive health.
Heroin has been a street drug at the Kenya coast since the 1980s where its use has spread from a few large towns to many smaller settlements, including some rural villages. The increasingly easy availability of heroin is linked to the 1980s tourist boom when Italian investors set up businesses with local partners. The Swahili community were particularly affected because they were in the forefront of the tourist industry and came into direct contact with Europeans requesting heroin [10].
As part of the Omari Project activities, MT and SB made assessments of the drug situation at the Kenya coast between 1995 and 1998 [8]. Until 1999, inhalation of vapour or 'chasing' 'brown sugar' was the dominant mode of use and the majority of heroin users at the Kenya coast were not injecting. 'White crest', a substance said by users to come from Thailand, started being mentioned in 1998 in Mombasa. 'White crest' cannot be chased but is easily injected. In late 1999 'brown sugar' disappeared from smaller coastal towns of Malindi and Watamu, and was replaced entirely by 'white crest'. Many chasers of 'brown sugar' became injectors of 'white crest'. The move to injecting was precipitated by the changes in the heroin supply that occurred in 1999. The UNDCP has been aware of heroin trafficking though the region. The publication by the UNDCP of trafficking routes from South Asia coincided with the decline in importance of such supply channels and the introduction of 'white crest' [11].
The ESRC study focused exclusively on Malindi, a town with a population of about 100,000. Since the 1960s Malindi has been a tourist resort. However, the town is an old Swahili city-state, already well established when Vasco da Gama visited 500 years ago en route to India. The original inhabitants of the area are Swahili fisher-people and traders. They have been joined by migrants from the rural hinterland and from the Kenya highlands, as well as by a significant minority of Europeans who own property in the area and are Kenyan residents.
In Malindi the switch to injecting heroin occurred in an area with high HIV rates. Sentinel surveillance carried out in antenatal clinics in the district showed a rate of 10% of attendees being HIV positive [12]. This figure was an average obtained from rural and urban areas within the district. In 2001 local health officials estimated that the HIV prevalence rate in the town was approximately 20%. It appears that the rate has remained at this high level, as about 20% of VCT clients undergoing HIV testing in the three centres operation in the District in July 2003 were positive (personal communication, VCT worker).
Esrc Study Methods
One objective of the ESRC study was to estimate the numbers of male and female users in the town in 2000 [13], while ethnographic fieldwork enabled a more in depth understanding of patterns of heroin use and the emerging sub-culture [14]. The methods used in the ESRC study to collect the findings reported here are summarised below:
Estimating user numbers
Users known to SB through her work with the Omari Project were asked to provide estimates of numbers of male and female users and how many were injectors. The lists of male users were cross-checked in interviews with other users to assess their reliability and to provide a source of data on which to base estimates of the number of male users residing in the town. The lists containing names of women users were the starting point for the snowball sample.
Snowball sampling has been used in the UK to contact drug users not known to the treatment services. The technique involves using each individual in the sample as a sampling node to generate the next subject until the sample is exhausted [15] and [16]. Snowball sampling has a number of potential drawbacks. For example, secret drug users who buy their supplies from different sources and use heroin alone could be missed from a sample generated from individuals who are part of a separate user network. Networks of heroin users might be formed along ethnic or class grounds, with members of one network having little or no knowledge of users in other networks.
In Malindi, Swahili people were concentrated in the old town area, but also lived in most neighbourhoods of the town. In addition, non-Swahili Kenyans were part of the same network of Swahili users living in the old town, and visited the area to purchase and use heroin and to meet other users. Men and women from various ethnic groups typically use heroin together, thereby overriding Swahili and Kenyan norms of gender and ethnic-based social segregation. However, a second network of European, mostly Italian users, exists, but was considered beyond the scope of the study. No attempt was made to include them in the estimates. In Malindi, where there appears to be one network of local users, a snowball sample was a useful means of contacting women users with a view to estimating their total number in the town.
Starting with two women already known to SB, female users were asked to list other women users. These were contacted and in turn asked to provide the names of other women users. They were also requested to participate in the study and assurances of confidentiality were made. Women were contacted all over the town until no new names appeared. In many cases this involved two 'generations' of individuals, and sometimes three. Throughout the research, SB updated her information on the identity of known women users, but during about two years of fieldwork few new names appeared.
(Participant)-Observation
Ethnographic research methods involving participant-observation that 'collect rich qualitative data' [17] enable researchers, who have gained the trust of groups of drug users, to observe their behaviour, hear about how they talk about drugs and join their social networks [18-21]. Observation is also a means of validating the accuracy of reported behaviour.
Between March 2000 and October 2001, participant observation and in-depth interviews amongst heroin users were carried out by SB, mostly during three main periods of fieldwork lasting between two to three months, but also during three shorter visits of between two to four weeks. Through initial contacts from amongst the users already known to SB or contacted with the assistance of a key informant, Ali, SB located 24 women users. These women were part of a bigger, predominantly male network of users. SB spent most of the time in the streets and alleyways where users congregate and also in the homes of about 20 female and male users, with whom she had built up rapport. She also conducted unstructured, in-depth one-to-one interviews with these users [13].
SB was able to observe heroin being smoked and injected and able to listen to discussion concerning raising the funds for and the purchasing of, heroin. She visited all areas of the town, but was most familiar with the old town area, where many of her initial key informants resided. Users acting as key informants were assured of confidentiality both for themselves and those that they named [22].
There was no option of working with drug agencies and organisations. However, SB already knew some heroin users through her connections with the Omari Project. At the time of the study reported here, the Omari Project had carried out some street work in the old town area and detoxified several local users at its small headquarters in a neighbouring town. Users were aware that SB was a member of the Omari Project and that a free residential service was opening shortly or had recently opened. Although SB explained that the study had no direct connection to the intervention activities of the Omari Project, users perceived her as somebody interested in their problems and who might be able to provide assistance. Simmons and Koester [23] report a similar situation in the US. Hence, SB was seen as a non-threatening, non-drug user and unlike Moore [18] in urban Australia, there was never any question that she should be a participant rather than a mere observer of heroin or any other drug use. As a woman in her forties SB was the same age as the parents of many of the users. Crucially, certain key users who enjoyed high status amongst their peers would vouch for her [24]. Those users who chose to talk to SB about their lives seemed to see her as a safe listener and keeper of secrets.
Estimates of User Numbers and Injectors
Our estimate of the total number of heroin users Malindi was 600 in the year 2000. This estimate was made after considering the lists produced by the male users for the old town area, and also taking into consideration the reported similar concentrations of users in three other neighbourhoods of the town. We traced or were informed of 26 women users in the town. The number of hidden and therefore uncounted female users is difficult to access, but is probably small. Hence, there were an estimated 30 women heroin users living in the town in 2000. During the two years of fieldwork, the number of about 25 known women users remained constant: although some female users died, moved away, went to prison or stopped using, they were replaced by others starting heroin use or moving into the area. Internationally, women form a minority of those in touch with drug services, and there are estimated to be far fewer women than men using heroin [25].
The assessments also included estimates of the proportion of heroin users who injected. The first male key informant approached by SB estimated that 80% of users in the town inject. Of the 15 named women users on the list provided by one women user, nine women were injecting and six were using by 'cocktail' (heroin mixed with cannabis) or 'joint' (heroin mixed with tobacco). The user who provided the list, a 'cocktail' user who had never injected, expressed shock at the high proportion of injectors on her list. Like most users, she perceived injecting to be more harmful than other modes of administration.
Hence, the percentage of injectors is difficult to estimate. According to the key informants in 2000, over half of users are injectors. However, such informants were long-term users who were likely to know other long-term users who had moved from smoking to injecting. New people are constantly being recruited to the ranks of heroin users and in this setting the vast majority of them do not start out as injectors. The estimated proportion of injectors, based on the multiple sources of data, was 50% in the year 2000.
Since 2000, there is nothing to indicate that the number of heroin users living in the town is decreasing. The price of a sachet of heroin has remained stable at KSh100 (approximately US$1.3). In 2003, outreach work carried out by the Omari Project located groups of new users concentrated in a suburb of the town where recent migrants congregate, an area which is also home to a prominent family of drug dealers. Most of the new users were heroin smokers, yet to make the transition to injecting. However, other users known to the Omari Project have moved to injecting within the last two years.
Heroin Injecting Culture
Language
SB found that the terminology used by members of the sub-culture of heroin users changes rapidly, so outsiders cannot readily gain entry to this group. As Ramos [26] found amongst Chicanos, ability to converse in this semi-secret language confers an insider status. The use of obscure terminology also assists heroin users in conducting their illegal and socially sanctioned activities within the midst of mainstream society. Heroin users in this Kenyan setting use a mixture of Swahili and English loan words to talk about injecting in particular and heroin use in general [10]. Most of these terms are slang and are not readily understood by Swahili-speakers who are not part of the heroin-using sub-culture. However, some terms such as 'junkie', 'shoot' or 'shooter' are understandable to drug users throughout the English-speaking world. Other terms are common to networks of heroin users within East Africa. Indeed, many words, such as tapeli ('scam') seem to originate from mainstream Tanzanian dialects of Swahili and have been diffused into Kenyan drug slang. Other terminology, such as kubwenga meaning 'to inject', or noma meaning 'a bad or dangerous incident' (such as being chased by the police), appears to be specific to the speech of heroin users at the Kenya Coast. However, such terms are likely to spread quickly into general street talk [27].
Injecting practice
Heroin users in Malindi have developed techniques and social protocols of injecting. In late 1999, when many users in the town moved from chasing 'brown sugar' to injecting 'white crest', one man was said by informants to have taught injection techniques for a fee. When interviewed he confirmed his 'teacher' role, adding that he now regretted being party to a change in heroin use that increased harm to users.
As users switched to injecting, many paid the fee and learnt how to inject themselves. Others remained dependent on 'doctors', users who inject others for a fee. Being a 'doctor' confers status and can be a source of easy money. Users who cannot inject themselves report having to raise the money to purchase heroin as well as an additional KSh40–50, representing about a 50% mark up on the price of the drug. Users report that a 'doctor's' services are paid for in cash and not in the form of a share of the heroin to be injected.
Injecting equipment
Needles and syringes are available to users in a few local pharmacies for between KSh5–10. Combined syringes and needles, designed for single use are not available and separate barrels and needles are purchased. Needles are large gauge 'blue' or 'green' needles. Their large size means that they are not ideal for injecting into small veins. Damage to veins, usually seen after several years of injection in the UK, was widespread among those who had been injecting for less than 6 months. This accelerates a move towards use of other injecting sites e.g. small veins in hands and feet, and sites in the groin where veins are in close proximity to an artery. These carry greater risks for the injector.
Most injectors use the same equipment to inject more than once, with some reporting using needles that have become rusty from being stored in damp hiding places. In addition, repeated use blunts the needle and eventually causes jamming of the syringe [28].
Injecting technique and sites
'White crest' used for injection is usually mixed with cold tap water. One or more sachets of one tenth of a gram of heroin are placed in the syringe. The required amount of water is drawn into the syringe and the solution shaken and examined for colour and to see that the 'white crest' has dissolved. If the user is injecting into the arm, a piece of string or rubber, or a belt or headscarf is tied round the upper arm as a tourniquet. Once the user finds a vein the needle is pushed in, the pump of the syringe is drawn back so that it fills with blood. The tourniquet is untied. Heroin is not always injected into the arms. Other injection sites amongst this group that I have observed are the legs, feet and backs of hands. Users also report injecting into the neck, near eyebrows, the groin area and the penis.
Users 'flush' a number of times. This procedure involves drawing blood back into the syringe and 'flushing' it back into the bloodstream. When asked, users have differing views on 'flushing'. Some say that it is best to flush several times, but that excessive flushing is damaging to the veins and smacks of desperation. Users will talk disparagingly of their friends who flush too many times while claiming that they themselves 'only flush two or three times'. Others claim that it is fine to flush as often as 'feels right'.
Getting, storing and disposing of needles and syringes
Needles and syringes are sold in Kenyan pharmacies for less than $0.10. However, some pharmacy salespeople refuse to sell injecting equipment to those they suspect are using illicit drugs. In Malindi, possession of used needles and syringes can lead to prosecution. Therefore, weighing up the relative risks of misplacing injecting equipment, another person borrowing it or using it because they mistake it for their own, as opposed to the danger of arrest for its possession, leads many injecting drug users (IDU) to decide not to carry injecting equipment on their person. Few IDU buy new equipment each time they use heroin, but conceal needles and syringes in locations where drugs are consumed. SB and MT have observed six or seven identical, unmarked syringes secreted under the eaves of a house near a major using location. SB has also watched one man wrapping his equipment in an old plastic bag and concealing it in weeds at the side of the path near his house. This policy may indeed be prudent. SB's key informant, Ali, explained how he was once chased by the police and subsequently arrested. The evidence against him was two sets of needles and syringes that the Police found concealed under a rock in the garden of his family home. The case went to court, but was dismissed when Ali argued that many people had access to the garden and there was no way of proving that they belonged to him. Of course, in other settings, blood tests or finger printing would have established ownership.
Used needles and syringes litter the floors of spaces frequented by injectors. One concerned old town resident collected a bag filled with discarded injecting equipment and left it on the doorstep of the main pharmacy supplying needles and syringes at that time. Nevertheless, many users display an almost complete lack of concern about the disposal of injecting equipment for which they have no further use. SB has seen them toss them into the grass beside a busy thoroughfare and throw them out of the windows of their homes.
'Partners'
Many users pair up with a 'partner' to raise cash, hang out and use together in a fashion similar to the strategies of Puerto Ricans living in the USA [28]. In Malindi, as elsewhere, for women who usually earn money through sex work, having a male partner can be a useful security measure. Mixed pairs are sometimes, but not always, sexual partners. Yet, the relationship is not primarily a sexual one, but is focussed on pairing up to support each other in the mutual aim of getting and using heroin, and same sex partners are numerous. Usually both partners are either smokers or injectors. If the pair are injectors, they may inject each other with the same or separate injecting equipment. Users all claim to, and appear to have, their own equipment.
Munira and her injecting
Munira was a young woman of about nineteen years of age. MT and SB had known her for several years, since she was first starting out as a heroin smoker. She earned money as a sex worker and was frequently involved in violent quarrels with other users and her family. Often, she raised money and used alone, although she sometimes paired up with an older woman. A detailed description of her injecting practice is provided elsewhere [30]. Below, SB's edited field notes illustrate the ways that her injecting was becoming out of control.
27.4.00
Ali says he saw Munira this morning waving a syringe around in the street and complaining that she has been sold whitewash. I saw Munira later and she confirmed the story about injecting whitewash. She explained that she did not bother to check the colour after adding water because she was in a hurry. When she got a vein, she pulled the syringe so got blood, but it would not push in. At home they gave her 100 Shillings to buy more heroin.
11.5.00
Even other users are particularly concerned about Munira. When we were talking to her, she developed a breathing problem and complaining of pain in her ribs. She is now resorting to injecting in the palm of her hand.
12.5.00
Munira followed Ali from near the premises of the main dealer, asking him to inject her in a ruined house nearby by the light of a candle. He told me the story and said that he refused.
15.5.00
Munira was almost caught injecting in the morning. She saw the policeman coming and stuck the syringe inside her blouse making her bleed.
The following year Munira was arrested on a theft charge and a used syringe was found hidden in her hair. The theft charges were eventually dropped against her when it emerged that she was merely collecting her fee for sex work from her client who was still sleeping. Nothing more was heard about drug charges.
Sharing
Although most Malindi users possess their own needles and syringes, sharing of injecting equipment occurs, as it is common amongst IDU in other settings [28,31,32]. Sharing can occur in a number of different ways, but does not appear to be perceived as a routine practice in Malindi.
Independently of each other, several users explained to SB the mechanism for sharing out a one-tenth measure of heroin. The powder is mixed with water and shaken in the syringe. The share, proportionate to the amount of money paid, is decanted into the plastic needle cover. It may then be drawn up into a second syringe. SB asked another informant, Ibrahim, how injectors share one sachet of heroin. He replied that with injecting as opposed to smoking, sharing one sachet is not done much. If it became necessary to share one sachet between two injectors, one sachet is put in the syringe, the water added, and then shares proportionate to the amount of money contributed are measured out into a second syringe belonging to the other person. Alternatively, the solution is put back into the original syringe once the other user has injected and rinsed out the equipment after use. The degree of safety or risk to HIV exposure will, of course, depend on a number of factors, including whether there are one or two sets of injecting equipment used, and if the equipment is new. These procedures are similar to the 'back loading' and 'frontloading' procedures described by injecting drug users in European and American settings [28,3].
Whilst these heroin users are aware that there may be health risks associated with sharing, they often seem to under-estimate the potential dangers and are vague about the illnesses that can be transmitted through injection sharing. For example SB asked Ibrahim why people did not like injecting together. He replied it was because of the illnesses they could catch from sharing equipment. When SB asked what these were, he mentioned 'AIDS, pneumonia, scabies, ringworm and problems inside the body'.
Although there has been no systematic campaign to inform them of the dangers of sharing equipment to inject illicit drugs, there have been a number of national AIDS prevention campaigns highlighting the possible dangers of acquiring HIV from used needles and syringes. Indeed, the ease by which injecting equipment can be purchased in private pharmacies is linked to HIV prevention initiatives. Over the years, the Omari Project, during counselling sessions, has also made efforts to point out the dangers of injecting.
Precautions taken by Ali and Elaine
When SB first met Ali he was the long-term sexual and using partner of Elaine. Elaine came from a wealthy background in the Kenya highlands and was better educated than most heroin users in the town. She had formerly owned a business, but had fallen on hard times. For the last couple of years she had lived with Ali, who had until recently been a successful dealer. Elaine described herself as a 'junkie' but took various measures in an attempt at discretion. Hence, for a period of time in 2000 she always injected into her legs, so as to avoid having track-marks on her arms. The trade off was sores and wounds on her ankles. By July 2000, Elaine was injecting into her arms, but with great difficulty. When SB asked Elaine about injecting practices, she confirmed what Ali had previously told me. When she was using with Ali in his family home one of them would agree to mark the syringe by burning the end, so that they knew whose was whose. The problem was that Ali's two brothers were also using in the family home and they could not be sure if they had used their (Ali's and Elaine's) syringes. However, she and Ali tried to hide them in places where they would not be found, and anyway, usually bought new ones everyday – others were put aside for standby in case the pharmacy was closed. She said that sometimes people asked to use her syringe after her – she told them she had AIDS, but they would reply that it did not matter. She said that people generally had their own syringes, but did not know if they marked them to distinguish them from others.
Social status and injecting
Injectors tend to be long-term users with high consumption levels. High consumption of heroin confers status, but only if used in a controlled fashion. Users should be able to raise or acquire money and drugs easily, have autonomy and avoid public displays of heroin use that indicate loss of self-control. Once the money to buy heroin is raised, it is preferable to have sufficient funds to buy one's own supplies without resorting to sharing with others, and also be able to inject oneself in a comfortable setting. High status users tend to inject themselves at home or at a friend's place. Some users are homeless and sleep on verandas, or in boats on the beach. Amongst homeless users, heroin use takes place in other locations, such as a derelict house or sometimes on the streets in the open. Injecting at home avoids inconvenience and reinforces the message that one has a home. Using alone denotes that one has sufficient funds, and is therefore status enhancing, while not having one's own injecting equipment and borrowing or stealing from another user denotes a lack of control over one's life and a lack of autonomy [13].
Many users who inject in semi-public settings are embarrassed by their injecting practices, perhaps sharing the widespread general aversion to needles and syringes or perhaps, because according to their own local knowledge, injecting denotes a deeper level of dependency than smoking. This embarrassment or shame concerning injecting practices is far from unique [28]. On the other hand, some users, like Munira and Ibrahim, seek out opportunities to inject in public settings or walk around the neighbourhood with a syringe sticking out of their arms. This public display is not status enhancing because it denotes a lack of self-control. Sitting or standing in the semi-conscious state (kuyoyoma) in a public place, is a practice that is looked down upon by many users.
By 2003 only one pharmacy in Malindi would sell needles and syringes to heroin users. There have been a number of deaths of Malindi users, all injectors or ex-injectors. Nevertheless, it can be expected that the majority were sufferers of AIDS. Although in Malindi, the subject of HIV and AIDS remains 'taboo' amongst users and non-users alike; those who shared injecting equipment with now deceased members of their network have expressed their fear and despair to members of the Omari Project that they are also HIV positive.
Discussion
The permutations of sharing injecting equipment with single or multiple partners within different venues, such as the home, public spaces or a shooting gallery have implications for the spread of HIV [3,31]. In Europe and North America much work on HIV awareness has been carried out, and most IDU are aware that sharing injecting equipment is a very risky activity. However, sharing continues to occur, particularly between sexual partners and when a user is suffering from heroin withdrawal. Nevertheless, the easy availability of new needles and syringes through needles exchanges has done much to reduce levels of sharing equipment. The situation in Kenya is different.
There is limited awareness of the dangers of HIV infection from sharing injection equipment, and limited access to new or clean equipment. The widespread injection of heroin is a recent development, yet injecting practices have acquired social significance. This new culture of injecting heroin and the developing protocols of sharing have implications for the transmission of HIV. While high status individuals sometimes help out other users who are in withdrawal by providing a few free puffs of heroin, they would not routinely allow their equipment to be used or to share injected heroin with others. This emphasis on individual use, as opposed to sharing heroin and injecting equipment, is an aspect of mainstream Kenyan life where individual effort and enjoyment of the fruits of personal endeavour are the dominant survival strategy [33].
Heroin use in Coastal Kenya should be recognised as part of the process of economic and cultural globalisation. Although large scale trafficking in heroin between South Asia and East Africa is a recent phenomenon dating from the late 1970s, there is a long history of contact between nations bordering the Indian Ocean. In Kenya, heroin use is spreading to smaller towns and even remote villages. Hence, 'drug abuse' is reported to be 'rampant' in Mbajumwali [34], a small village on Pate Island within Lamu District on the North Kenya coast. This newspaper report confirms that heroin is easily available in Mbajumwali and the nearby bigger village of Kizingtini.
In other settings, the easy availability of new needles and syringes through needle exchanges has done much to reduce levels of sharing equipment, but this opportunity has not been available to injectors in Kenya. Studies from other parts of the world demonstrate that needle exchanges can reduce the transmission of HIV and other blood borne diseases [35-37], but that needle exchanges are most effective when part of 'an integrated set of diverse preventive measures' [38] as has been applied in Australia. Indeed, some studies in American and European cities demonstrate that needle and syringe exchanges have not prevented rises in HIV and Hepatitis B and C infections; such rises occurring among drug injectors in Vancouver [39], Montreal [40] and Amsterdam [41]. The reasons for the failure in prevention of HIV and Hepatitis B and C were the dilution of the impact of needle exchange schemes, due to an inadequate volume of syringes supplied (especially where cocaine injection was widespread as in Vancouver), insufficient focus on the risks of sharing injection paraphernalia, little impact on sexual behaviour and a non-interventionist ethos.
Developing needle exchange services in Kenya carries a clear responsibility to learn from such research. The findings of these studies indicate that the provision of adequate supplies of injecting equipment must be coupled with an interventionist ethos, whereby risk behaviour is reviewed regularly with injectors and strategies are developed to minimise these. This challenge is both exciting and immense.
Conclusion
Towards a Regional Response
Odek-Ogunde et al [5] found 44.9 % of heroin users in Nairobi had been or were injectors. In addition, 'white crest' is widely available in Nairobi and injection techniques are similar to those observed at the Coast (Odek-Ogunde, personal communication). Kenyan heroin users and development workers report that heroin is even more widely used and easily available in Nairobi and Tanzanian cities such as Zanzibar, Arusha and Dar es Salaam, than at the Kenyan Coast. The recent study from Dar es Salaam, Tanzania [7] indicates that injecting drug use is widespread in the Tanzanian capital. Similarly, the 2001 assessment of drug use in Tanzanian towns [6] found injecting drug use to be a matter of serious concern and advocated the setting up of needle and syringe exchange programmes. However, injecting drug use in Tanzania extends beyond the areas assessed in the 2001 situational analysis. For example, anecdotal reports by the Zanzibar Youth Forum (personal communication) indicate that rates of injecting are higher on the more remote island of Pemba than in Zanzibar town (the islands of Pemba and Unguja comprise Zanzibar, which is also the name of the main town on the island of Unguja).
Heroin users at the Kenya Coast, as elsewhere within the region, lack information about the dangers of injecting, the risks of sharing needles and syringes and unprotected sex. Odek-Ogunde has called for harm reduction measures and notes that 'the presence of IDUs in Kenya creates a three fold risk: the escalation of injection drug use, the potential of exacerbating the HIV/AIDS situation, and the creation of a source for other drug-related damage' [42].
Coastal Kenyan heroin injectors share injecting equipment and have sex with each other as well as with non-users. Their risky behaviour takes place in a setting where about 20% of the general population are estimated to be HIV positive [13]. The need for harm reduction initiatives amongst this high-risk group is clear. The findings of the study in Nairobi, which show that 1 in 2 injectors interviewed is HIV Positive, demand an urgent response [5]. Heroin users lack information about the dangers of injecting, of sharing needles and syringes, and of unprotected sex. Most East African towns have a lively sex industry and indeed the vast majority of women users work in this field.
Best practice from around the world demonstrates that HIV prevention through harm reduction measures should continue to target 'high risk groups' even when the epidemic has moved into the general population [2,4].
In November 2003, the Omari Project obtained funding from the UK Community Fund to open a drop-in centre in Malindi where harm reduction strategies, including needle exchanges will be pioneered. This offers an opportunity to curtail the transmission of HIV and other infections such as Hepatitis C, the prevalence of which is unknown in Kenya, although data from Nairobi indicates that both are very prevalent. The development of this intervention also offers the opportunity to learn from studies of needle exchange 'failures' in Europe and North America and to avoid repeating their unwitting mistakes.
The UK Department for International Development (DFID) has expressed interest in supporting such initiatives and extending them to other East African cities, and is providing funding to the Omari Project to assess injection drug use in Mombasa, with a view to developing appropriate harm reduction services, including a needle exchange. As in most parts of the world [29], functional needle exchange programmes will require careful advocacy work if the police and community are not to oppose their operation. The ESRC study in collaboration with the Omari Project, reported above, has already held a series of workshops with health workers and key community members in Malindi, where the dangers of injecting drug use and its relationship with HIV have been highlighted [14].
Acknowledgements
We thank the ESRC for funding this research (Grant no ROOO 23 8392). We also thank the men and women users who participated in the study. The Omari Project and Bristol Drugs Project thank the UK Community Fund for their continuing support. DFID and their managing agents Futures Group Europe are also thanked for their support in the pioneering of harm reduction approaches in East Africa.
==== Refs
Mwenesi HA Rapid assessment of Drug Use in Kenya Bulletin on Narcotics 1996 XLVIII 265 278
Ball AL Rana S Dehne KL HIV prevention amongst injecting drug users: responses in developing and transitional countries Public Health Report 1998 113 170 181
Des Jarlais DC Dehne K Casabona J HIV surveillance among injecting drug users AIDS 2001 15 S13 S22 11421178 10.1097/00002030-200104003-00003
Stimson G Choopanya K Stimson G, Des Jarlais DC Global Perspectives on Drug Injecting Drug Injecting and HIV Infection 1998 Ball A London: UCL Press
Odek-Ogunde M Lore W Owiti FR Risky behaviours among injecting drug users in Kenya Proceedings of 14th International Conference on the Reduction of Drug Related Harm 2003 Chiang Mai, Thailand
Kilonzo GP Mbwambo JK Kaaya SF Hogan NM Rapid Situational Assessment for Drug Demand Reduction in Tanzania UNDCP 2001
Muhondwa E Mpembeni R Mayunga F An assessment of the treatment needs of drug users un Dar es Salaam Report by Christ Compassion in Action prepared for Save the Children UK, Tanzania Programme 2002
Beckerleg S Telfer M Kibwana Sizi A Private struggles, public support: rehabilitating heroin users in Kenya Drugs: Education, Prevention and Policy 1996 3
Beckerleg S Counselling Kenyan Heroin Users: Cross Cultural Motivation? Health Education 2001 101
Beckerleg S Brown Sugar or Friday Prayers: Youth choices and community building in Coastal Kenya African Affairs 1995 94 23 38
UNDCP on The Eastern African Drug Beat Importance of Data in Control of Illicit Drug Trafficking 2000 2 13 14
Republic of Kenya on AIDS in Kenya 2001 Sixth Ministry of Health
Beckerleg S Lewando Hundt G (a) The characteristics and recent growth of heroin injecting in a Kenyan coastal town Addiction Research and Theory Forthcoming
Beckerleg S Lewando Hundt G L Hume, J Mulcock (b) Reflections on Fieldwork amongst Heroin Users in a Small town in Kenya Awkward Spaces, Productive Spaces: Fieldwork in Difficult Conditions Columbia University Press Forthcoming
Griffiths P Gossop M Powis B Strang J Reaching hidden populations of drug users by privileged access interviewers: methodological and practical issues Addiction 1993 88 1617 1626 8130701
Van Meter KM Lambert E Methodological and Design Issues: Techniques for Assessing the Representatives of Snowball Samples The Collection and Interpretation of data from hidden populations 1990 NIDA: Research Monograph Series 98: Rockville: NIDA
Power R A model for qualitative action research amongst illicit drug users Addiction Research 1995 3 165 181
Moore D Ethnography and Illicit Drug Use: dispatches from an anthropologist in the 'Field' Addiction Research 1993 1
Power R Rapid Assessment of the Drug-injecting situation at Hanoi and Ho Chi Minh City, Viet Nam Bulletin on Narcotics 1996 XLVIII 35 52 9839034
Power R Reflections on participant observation in drugs research Addiction Research and Theory 2001 9 325 337
Kaplan C Verbraeck H Where have all the fieldnotes gone? The changing nature and politics of drugs ethnography in The Netherlands Addiction Research and Theory 2001 9 229 323
Nuffield Council of Bioethics on The ethics of research related to healthcare in developing countries 2002
Simmons J Koester K Hidden injuries of research on social suffering among drug users Practicing Anthropology 2003 25 53 57
Wiebel W Lambert E Identifying and Gaining Access to Hidden Populations The collection and interpretation of data from hidden populations 1990 NIDA, Research Monograph Series 98: Rockville: NIDA
Harwood H Fountain D Carothers S Gerstein D Johnson R Stevens SJ, Wexler HK Gender differences in economic impacts of clients before, during and after substance abuse treatment Women and Substance Abuse 1998 London: The Haworth Press
Ramos R Lambert E Chicano Intravenous Drug Users The collection and interpretation of data from hidden populations 1990 NIDA, Research Monograph Series 98: Rockville: NIDA
Bourgois P Search of Respect: Selling Crack in El Barrio 1995 Cambridge: Cambridge University Press
Lockley P Counselling Heroin and Other Drug Users 1995 London: Free Association Books
Santelices C Singer M Nicolaysen AM Risky and precarious dependencies of Puerto Rican "IDUS" in El Barrio: an ethnographic glimpse Practicing Anthropology 2003 25 23 27
Beckerleg S Coomber R, South N Living with Heroin at the Kenya Coast Drugs in the Global Context: comparative perspectives on cultures and controls Free Association Press Forthcoming
Power R Jones S Kearns G Ward J An ethnography of risk management amongst illicit drug injectors and its implications for the development of community-based interventions Sociology of Health and Illness 1996 18 86 106 10.1111/1467-9566.ep10934419
Hughes RA Drug injectors and the cleaning of needles and syringes European Addiction Research 2000 6 20 30 10729739 10.1159/000019005
UNODCCP on The Drug Nexus in Africa 1999 UN Office for Drug Control & Crime Prevention Monographs, Vienna
Ndurya M Red Alert on Drug Deaths Coast Express 2003 1 7
Lurie P Drucker E An opportunity lost: HIV infections associated with lack of a needle-exchange program in the USA Lancet 349 604 608 1997, March 1 9057732 10.1016/S0140-6736(96)05439-6
Lurie P Reingold A School of Public Health on The public health impact of needle exchange programs in the United States and abroad 1993 Berkley, CA
Ksobiech A meta analysis of needle sharing, lending and borrowing behaviours of nep attenders AIDS: Education and Prevention 2003 15
Bastos F Stimson GV Telles P Barcellos C Stimson G, Des Jarlais DC, Ball A Cities responding to HIV-1 epidemics among injecting drug users Drug Injecting and HIV Infection 1998 London: UCL Press 149 167
Schechter MT Strathdee SA Cornelisse PJA Currie S Patrick DM Rekart ML O'Shaughnessy MV Do needle exchange programmes increase the spread of HIV among injection drug users? An investigation of the Vancouver outbreak AIDS 1999 13
Bruneaus J Langendam MW Notenboom J Coutinho RA High rates of HIV infection among injection drug users participating in needle exchange programmes in Montreal: results of a cohort study American Journal of Epidemiology 1997 146
Van Ameijden EJC Bruneau J Lamothe F Franco E Lachance N Desy M Soto J Vincelette J Continuing injecting risk behaviour: results from the Amsterdam cohort study of drug users Addiction 1999 94
Odek-Ogunde M Harm reduction as alternative strategy for prevention of drug abuse and HIV/AIDS in Kenya Proceedings of Third Annual Meeting of Global Research Network Meeting on HIV Prevention in Drug-Using populations 2000 Durban, South Africa 113 117
|
16122382
|
PMC1236949
|
CC BY
|
2021-01-04 16:36:49
|
no
|
Harm Reduct J. 2005 Aug 25; 2:12
|
utf-8
|
Harm Reduct J
| 2,005 |
10.1186/1477-7517-2-12
|
oa_comm
|
==== Front
Harm Reduct JHarm Reduction Journal1477-7517BioMed Central London 1477-7517-2-121612238210.1186/1477-7517-2-12CommentaryThe rise of injecting drug use in east Africa: a case study from Kenya Beckerleg Susan [email protected] Maggie [email protected] Gillian Lewando [email protected] Public and Environmental Health Research Unit, London School of Hygiene & Tropical Medicine, London, UK2 Bristol Drugs Project, 11 Brunswick Square, Bristol, UK3 Social Sciences in Health, University of Warwick, UK2005 25 8 2005 2 12 12 11 7 2004 25 8 2005 Copyright © 2005 Beckerleg et al; licensee BioMed Central Ltd.2005Beckerleg 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.
Studies on injecting drug use in East Africa are reviewed. The existingstudies document the spread of heroin injection in Kenya and Tanzania, both countries where HIV rates are high. No data from Uganda on injecting drug use was found by the authors. A case study of the growth of heroin injection in a Kenyan coastal town is presented. The need for needle-exchange programmes and other prevention services is discussed.
==== Body
Background
Although bearing the brunt of the AIDS epidemic, Africa has long been considered largely free of injection drug use. Notwithstanding the assessments of the UN International Drug Control Programme [1] 1, international organisations have been slow to recognise either the spread of heroin use in Kenya or the existence of injection drug use. The largely unheeded spread of injection drug use in East Africa has wide implications for public health in the region. Injection drug users (IDU) are a 'high risk' or 'core group' for HIV infection. Many IDU share needles and syringes as well as having unprotected sex, and have been identified as a 'bridging population', speeding the spread of HIV to the general population [2,3] and [4].
Heroin injection now appears to be occurring in most large towns of Kenya and Tanzania. A study of 336 heroin users in Nairobi, Kenya found that 44.9% were, or had been, injectors [5]. Of 101 current injectors, 52.5% were HIV positive. This compares with a 13.5% prevalence rate among heroin users who had never injected. Hepatitis C prevalence also varied dramatically, from 61.4% among current injectors to 3.8% for those who had never injected. A similar study for Mombasa, Kenya's second city and main port, has been planned, but at the time of writing (November 2003) is yet to be carried out. However the UNODC and WHO are carrying out research in 2003 to establish links between HIV and drugs, with a focus on injecting, at the Kenya Coast.
Recent assessments in neighbouring Tanzania have found heroin injecting to be spreading throughout the country. Hence, a rapid situation assessment carried out in five Tanzanian towns [6] found heroin to be a major concern in Arusha, Dar es Salaam and Zanzibar, to be emerging as problem in Mwanza, but not in Mbeya. Injection drug use was reported in all the study sites where heroin was in use. Similarly, a study of 624 young multi-drug (alcohol, cannabis, tobacco, heroin, Valium, khat) users in Dar es Salaam found that 75% of the sample were using heroin, and that 114 (18.3%) of the sample reported injecting drugs [7]. As many of the substances used by the 624 people interviewed are not usually injected, the percentage of heroin users injecting in Dar es Salaam will be considerably higher than is indicated by these data which are not disaggregated by substance.
Much less is known about injecting drug use in Uganda. Indeed, the UNODC covering Eastern and Southern Africa reports that there have been no drug assessments carried out in Uganda.
Discussion
Case Study: The Kenya Coast
The case study draws together information collected by SB and MT from 1995 to 2003 as part of their work with the Omari Project, a Kenyan organisation pioneering the adaptation of international best practice treatment models for heroin users to the local setting of Coastal Kenya [8] and [9]. In addition, SB collected data on heroin users and injecting patterns between 2000 and 2001, as part of a larger ESRC funded study of women heroin users and their reproductive health.
Heroin has been a street drug at the Kenya coast since the 1980s where its use has spread from a few large towns to many smaller settlements, including some rural villages. The increasingly easy availability of heroin is linked to the 1980s tourist boom when Italian investors set up businesses with local partners. The Swahili community were particularly affected because they were in the forefront of the tourist industry and came into direct contact with Europeans requesting heroin [10].
As part of the Omari Project activities, MT and SB made assessments of the drug situation at the Kenya coast between 1995 and 1998 [8]. Until 1999, inhalation of vapour or 'chasing' 'brown sugar' was the dominant mode of use and the majority of heroin users at the Kenya coast were not injecting. 'White crest', a substance said by users to come from Thailand, started being mentioned in 1998 in Mombasa. 'White crest' cannot be chased but is easily injected. In late 1999 'brown sugar' disappeared from smaller coastal towns of Malindi and Watamu, and was replaced entirely by 'white crest'. Many chasers of 'brown sugar' became injectors of 'white crest'. The move to injecting was precipitated by the changes in the heroin supply that occurred in 1999. The UNDCP has been aware of heroin trafficking though the region. The publication by the UNDCP of trafficking routes from South Asia coincided with the decline in importance of such supply channels and the introduction of 'white crest' [11].
The ESRC study focused exclusively on Malindi, a town with a population of about 100,000. Since the 1960s Malindi has been a tourist resort. However, the town is an old Swahili city-state, already well established when Vasco da Gama visited 500 years ago en route to India. The original inhabitants of the area are Swahili fisher-people and traders. They have been joined by migrants from the rural hinterland and from the Kenya highlands, as well as by a significant minority of Europeans who own property in the area and are Kenyan residents.
In Malindi the switch to injecting heroin occurred in an area with high HIV rates. Sentinel surveillance carried out in antenatal clinics in the district showed a rate of 10% of attendees being HIV positive [12]. This figure was an average obtained from rural and urban areas within the district. In 2001 local health officials estimated that the HIV prevalence rate in the town was approximately 20%. It appears that the rate has remained at this high level, as about 20% of VCT clients undergoing HIV testing in the three centres operation in the District in July 2003 were positive (personal communication, VCT worker).
Esrc Study Methods
One objective of the ESRC study was to estimate the numbers of male and female users in the town in 2000 [13], while ethnographic fieldwork enabled a more in depth understanding of patterns of heroin use and the emerging sub-culture [14]. The methods used in the ESRC study to collect the findings reported here are summarised below:
Estimating user numbers
Users known to SB through her work with the Omari Project were asked to provide estimates of numbers of male and female users and how many were injectors. The lists of male users were cross-checked in interviews with other users to assess their reliability and to provide a source of data on which to base estimates of the number of male users residing in the town. The lists containing names of women users were the starting point for the snowball sample.
Snowball sampling has been used in the UK to contact drug users not known to the treatment services. The technique involves using each individual in the sample as a sampling node to generate the next subject until the sample is exhausted [15] and [16]. Snowball sampling has a number of potential drawbacks. For example, secret drug users who buy their supplies from different sources and use heroin alone could be missed from a sample generated from individuals who are part of a separate user network. Networks of heroin users might be formed along ethnic or class grounds, with members of one network having little or no knowledge of users in other networks.
In Malindi, Swahili people were concentrated in the old town area, but also lived in most neighbourhoods of the town. In addition, non-Swahili Kenyans were part of the same network of Swahili users living in the old town, and visited the area to purchase and use heroin and to meet other users. Men and women from various ethnic groups typically use heroin together, thereby overriding Swahili and Kenyan norms of gender and ethnic-based social segregation. However, a second network of European, mostly Italian users, exists, but was considered beyond the scope of the study. No attempt was made to include them in the estimates. In Malindi, where there appears to be one network of local users, a snowball sample was a useful means of contacting women users with a view to estimating their total number in the town.
Starting with two women already known to SB, female users were asked to list other women users. These were contacted and in turn asked to provide the names of other women users. They were also requested to participate in the study and assurances of confidentiality were made. Women were contacted all over the town until no new names appeared. In many cases this involved two 'generations' of individuals, and sometimes three. Throughout the research, SB updated her information on the identity of known women users, but during about two years of fieldwork few new names appeared.
(Participant)-Observation
Ethnographic research methods involving participant-observation that 'collect rich qualitative data' [17] enable researchers, who have gained the trust of groups of drug users, to observe their behaviour, hear about how they talk about drugs and join their social networks [18-21]. Observation is also a means of validating the accuracy of reported behaviour.
Between March 2000 and October 2001, participant observation and in-depth interviews amongst heroin users were carried out by SB, mostly during three main periods of fieldwork lasting between two to three months, but also during three shorter visits of between two to four weeks. Through initial contacts from amongst the users already known to SB or contacted with the assistance of a key informant, Ali, SB located 24 women users. These women were part of a bigger, predominantly male network of users. SB spent most of the time in the streets and alleyways where users congregate and also in the homes of about 20 female and male users, with whom she had built up rapport. She also conducted unstructured, in-depth one-to-one interviews with these users [13].
SB was able to observe heroin being smoked and injected and able to listen to discussion concerning raising the funds for and the purchasing of, heroin. She visited all areas of the town, but was most familiar with the old town area, where many of her initial key informants resided. Users acting as key informants were assured of confidentiality both for themselves and those that they named [22].
There was no option of working with drug agencies and organisations. However, SB already knew some heroin users through her connections with the Omari Project. At the time of the study reported here, the Omari Project had carried out some street work in the old town area and detoxified several local users at its small headquarters in a neighbouring town. Users were aware that SB was a member of the Omari Project and that a free residential service was opening shortly or had recently opened. Although SB explained that the study had no direct connection to the intervention activities of the Omari Project, users perceived her as somebody interested in their problems and who might be able to provide assistance. Simmons and Koester [23] report a similar situation in the US. Hence, SB was seen as a non-threatening, non-drug user and unlike Moore [18] in urban Australia, there was never any question that she should be a participant rather than a mere observer of heroin or any other drug use. As a woman in her forties SB was the same age as the parents of many of the users. Crucially, certain key users who enjoyed high status amongst their peers would vouch for her [24]. Those users who chose to talk to SB about their lives seemed to see her as a safe listener and keeper of secrets.
Estimates of User Numbers and Injectors
Our estimate of the total number of heroin users Malindi was 600 in the year 2000. This estimate was made after considering the lists produced by the male users for the old town area, and also taking into consideration the reported similar concentrations of users in three other neighbourhoods of the town. We traced or were informed of 26 women users in the town. The number of hidden and therefore uncounted female users is difficult to access, but is probably small. Hence, there were an estimated 30 women heroin users living in the town in 2000. During the two years of fieldwork, the number of about 25 known women users remained constant: although some female users died, moved away, went to prison or stopped using, they were replaced by others starting heroin use or moving into the area. Internationally, women form a minority of those in touch with drug services, and there are estimated to be far fewer women than men using heroin [25].
The assessments also included estimates of the proportion of heroin users who injected. The first male key informant approached by SB estimated that 80% of users in the town inject. Of the 15 named women users on the list provided by one women user, nine women were injecting and six were using by 'cocktail' (heroin mixed with cannabis) or 'joint' (heroin mixed with tobacco). The user who provided the list, a 'cocktail' user who had never injected, expressed shock at the high proportion of injectors on her list. Like most users, she perceived injecting to be more harmful than other modes of administration.
Hence, the percentage of injectors is difficult to estimate. According to the key informants in 2000, over half of users are injectors. However, such informants were long-term users who were likely to know other long-term users who had moved from smoking to injecting. New people are constantly being recruited to the ranks of heroin users and in this setting the vast majority of them do not start out as injectors. The estimated proportion of injectors, based on the multiple sources of data, was 50% in the year 2000.
Since 2000, there is nothing to indicate that the number of heroin users living in the town is decreasing. The price of a sachet of heroin has remained stable at KSh100 (approximately US$1.3). In 2003, outreach work carried out by the Omari Project located groups of new users concentrated in a suburb of the town where recent migrants congregate, an area which is also home to a prominent family of drug dealers. Most of the new users were heroin smokers, yet to make the transition to injecting. However, other users known to the Omari Project have moved to injecting within the last two years.
Heroin Injecting Culture
Language
SB found that the terminology used by members of the sub-culture of heroin users changes rapidly, so outsiders cannot readily gain entry to this group. As Ramos [26] found amongst Chicanos, ability to converse in this semi-secret language confers an insider status. The use of obscure terminology also assists heroin users in conducting their illegal and socially sanctioned activities within the midst of mainstream society. Heroin users in this Kenyan setting use a mixture of Swahili and English loan words to talk about injecting in particular and heroin use in general [10]. Most of these terms are slang and are not readily understood by Swahili-speakers who are not part of the heroin-using sub-culture. However, some terms such as 'junkie', 'shoot' or 'shooter' are understandable to drug users throughout the English-speaking world. Other terms are common to networks of heroin users within East Africa. Indeed, many words, such as tapeli ('scam') seem to originate from mainstream Tanzanian dialects of Swahili and have been diffused into Kenyan drug slang. Other terminology, such as kubwenga meaning 'to inject', or noma meaning 'a bad or dangerous incident' (such as being chased by the police), appears to be specific to the speech of heroin users at the Kenya Coast. However, such terms are likely to spread quickly into general street talk [27].
Injecting practice
Heroin users in Malindi have developed techniques and social protocols of injecting. In late 1999, when many users in the town moved from chasing 'brown sugar' to injecting 'white crest', one man was said by informants to have taught injection techniques for a fee. When interviewed he confirmed his 'teacher' role, adding that he now regretted being party to a change in heroin use that increased harm to users.
As users switched to injecting, many paid the fee and learnt how to inject themselves. Others remained dependent on 'doctors', users who inject others for a fee. Being a 'doctor' confers status and can be a source of easy money. Users who cannot inject themselves report having to raise the money to purchase heroin as well as an additional KSh40–50, representing about a 50% mark up on the price of the drug. Users report that a 'doctor's' services are paid for in cash and not in the form of a share of the heroin to be injected.
Injecting equipment
Needles and syringes are available to users in a few local pharmacies for between KSh5–10. Combined syringes and needles, designed for single use are not available and separate barrels and needles are purchased. Needles are large gauge 'blue' or 'green' needles. Their large size means that they are not ideal for injecting into small veins. Damage to veins, usually seen after several years of injection in the UK, was widespread among those who had been injecting for less than 6 months. This accelerates a move towards use of other injecting sites e.g. small veins in hands and feet, and sites in the groin where veins are in close proximity to an artery. These carry greater risks for the injector.
Most injectors use the same equipment to inject more than once, with some reporting using needles that have become rusty from being stored in damp hiding places. In addition, repeated use blunts the needle and eventually causes jamming of the syringe [28].
Injecting technique and sites
'White crest' used for injection is usually mixed with cold tap water. One or more sachets of one tenth of a gram of heroin are placed in the syringe. The required amount of water is drawn into the syringe and the solution shaken and examined for colour and to see that the 'white crest' has dissolved. If the user is injecting into the arm, a piece of string or rubber, or a belt or headscarf is tied round the upper arm as a tourniquet. Once the user finds a vein the needle is pushed in, the pump of the syringe is drawn back so that it fills with blood. The tourniquet is untied. Heroin is not always injected into the arms. Other injection sites amongst this group that I have observed are the legs, feet and backs of hands. Users also report injecting into the neck, near eyebrows, the groin area and the penis.
Users 'flush' a number of times. This procedure involves drawing blood back into the syringe and 'flushing' it back into the bloodstream. When asked, users have differing views on 'flushing'. Some say that it is best to flush several times, but that excessive flushing is damaging to the veins and smacks of desperation. Users will talk disparagingly of their friends who flush too many times while claiming that they themselves 'only flush two or three times'. Others claim that it is fine to flush as often as 'feels right'.
Getting, storing and disposing of needles and syringes
Needles and syringes are sold in Kenyan pharmacies for less than $0.10. However, some pharmacy salespeople refuse to sell injecting equipment to those they suspect are using illicit drugs. In Malindi, possession of used needles and syringes can lead to prosecution. Therefore, weighing up the relative risks of misplacing injecting equipment, another person borrowing it or using it because they mistake it for their own, as opposed to the danger of arrest for its possession, leads many injecting drug users (IDU) to decide not to carry injecting equipment on their person. Few IDU buy new equipment each time they use heroin, but conceal needles and syringes in locations where drugs are consumed. SB and MT have observed six or seven identical, unmarked syringes secreted under the eaves of a house near a major using location. SB has also watched one man wrapping his equipment in an old plastic bag and concealing it in weeds at the side of the path near his house. This policy may indeed be prudent. SB's key informant, Ali, explained how he was once chased by the police and subsequently arrested. The evidence against him was two sets of needles and syringes that the Police found concealed under a rock in the garden of his family home. The case went to court, but was dismissed when Ali argued that many people had access to the garden and there was no way of proving that they belonged to him. Of course, in other settings, blood tests or finger printing would have established ownership.
Used needles and syringes litter the floors of spaces frequented by injectors. One concerned old town resident collected a bag filled with discarded injecting equipment and left it on the doorstep of the main pharmacy supplying needles and syringes at that time. Nevertheless, many users display an almost complete lack of concern about the disposal of injecting equipment for which they have no further use. SB has seen them toss them into the grass beside a busy thoroughfare and throw them out of the windows of their homes.
'Partners'
Many users pair up with a 'partner' to raise cash, hang out and use together in a fashion similar to the strategies of Puerto Ricans living in the USA [28]. In Malindi, as elsewhere, for women who usually earn money through sex work, having a male partner can be a useful security measure. Mixed pairs are sometimes, but not always, sexual partners. Yet, the relationship is not primarily a sexual one, but is focussed on pairing up to support each other in the mutual aim of getting and using heroin, and same sex partners are numerous. Usually both partners are either smokers or injectors. If the pair are injectors, they may inject each other with the same or separate injecting equipment. Users all claim to, and appear to have, their own equipment.
Munira and her injecting
Munira was a young woman of about nineteen years of age. MT and SB had known her for several years, since she was first starting out as a heroin smoker. She earned money as a sex worker and was frequently involved in violent quarrels with other users and her family. Often, she raised money and used alone, although she sometimes paired up with an older woman. A detailed description of her injecting practice is provided elsewhere [30]. Below, SB's edited field notes illustrate the ways that her injecting was becoming out of control.
27.4.00
Ali says he saw Munira this morning waving a syringe around in the street and complaining that she has been sold whitewash. I saw Munira later and she confirmed the story about injecting whitewash. She explained that she did not bother to check the colour after adding water because she was in a hurry. When she got a vein, she pulled the syringe so got blood, but it would not push in. At home they gave her 100 Shillings to buy more heroin.
11.5.00
Even other users are particularly concerned about Munira. When we were talking to her, she developed a breathing problem and complaining of pain in her ribs. She is now resorting to injecting in the palm of her hand.
12.5.00
Munira followed Ali from near the premises of the main dealer, asking him to inject her in a ruined house nearby by the light of a candle. He told me the story and said that he refused.
15.5.00
Munira was almost caught injecting in the morning. She saw the policeman coming and stuck the syringe inside her blouse making her bleed.
The following year Munira was arrested on a theft charge and a used syringe was found hidden in her hair. The theft charges were eventually dropped against her when it emerged that she was merely collecting her fee for sex work from her client who was still sleeping. Nothing more was heard about drug charges.
Sharing
Although most Malindi users possess their own needles and syringes, sharing of injecting equipment occurs, as it is common amongst IDU in other settings [28,31,32]. Sharing can occur in a number of different ways, but does not appear to be perceived as a routine practice in Malindi.
Independently of each other, several users explained to SB the mechanism for sharing out a one-tenth measure of heroin. The powder is mixed with water and shaken in the syringe. The share, proportionate to the amount of money paid, is decanted into the plastic needle cover. It may then be drawn up into a second syringe. SB asked another informant, Ibrahim, how injectors share one sachet of heroin. He replied that with injecting as opposed to smoking, sharing one sachet is not done much. If it became necessary to share one sachet between two injectors, one sachet is put in the syringe, the water added, and then shares proportionate to the amount of money contributed are measured out into a second syringe belonging to the other person. Alternatively, the solution is put back into the original syringe once the other user has injected and rinsed out the equipment after use. The degree of safety or risk to HIV exposure will, of course, depend on a number of factors, including whether there are one or two sets of injecting equipment used, and if the equipment is new. These procedures are similar to the 'back loading' and 'frontloading' procedures described by injecting drug users in European and American settings [28,3].
Whilst these heroin users are aware that there may be health risks associated with sharing, they often seem to under-estimate the potential dangers and are vague about the illnesses that can be transmitted through injection sharing. For example SB asked Ibrahim why people did not like injecting together. He replied it was because of the illnesses they could catch from sharing equipment. When SB asked what these were, he mentioned 'AIDS, pneumonia, scabies, ringworm and problems inside the body'.
Although there has been no systematic campaign to inform them of the dangers of sharing equipment to inject illicit drugs, there have been a number of national AIDS prevention campaigns highlighting the possible dangers of acquiring HIV from used needles and syringes. Indeed, the ease by which injecting equipment can be purchased in private pharmacies is linked to HIV prevention initiatives. Over the years, the Omari Project, during counselling sessions, has also made efforts to point out the dangers of injecting.
Precautions taken by Ali and Elaine
When SB first met Ali he was the long-term sexual and using partner of Elaine. Elaine came from a wealthy background in the Kenya highlands and was better educated than most heroin users in the town. She had formerly owned a business, but had fallen on hard times. For the last couple of years she had lived with Ali, who had until recently been a successful dealer. Elaine described herself as a 'junkie' but took various measures in an attempt at discretion. Hence, for a period of time in 2000 she always injected into her legs, so as to avoid having track-marks on her arms. The trade off was sores and wounds on her ankles. By July 2000, Elaine was injecting into her arms, but with great difficulty. When SB asked Elaine about injecting practices, she confirmed what Ali had previously told me. When she was using with Ali in his family home one of them would agree to mark the syringe by burning the end, so that they knew whose was whose. The problem was that Ali's two brothers were also using in the family home and they could not be sure if they had used their (Ali's and Elaine's) syringes. However, she and Ali tried to hide them in places where they would not be found, and anyway, usually bought new ones everyday – others were put aside for standby in case the pharmacy was closed. She said that sometimes people asked to use her syringe after her – she told them she had AIDS, but they would reply that it did not matter. She said that people generally had their own syringes, but did not know if they marked them to distinguish them from others.
Social status and injecting
Injectors tend to be long-term users with high consumption levels. High consumption of heroin confers status, but only if used in a controlled fashion. Users should be able to raise or acquire money and drugs easily, have autonomy and avoid public displays of heroin use that indicate loss of self-control. Once the money to buy heroin is raised, it is preferable to have sufficient funds to buy one's own supplies without resorting to sharing with others, and also be able to inject oneself in a comfortable setting. High status users tend to inject themselves at home or at a friend's place. Some users are homeless and sleep on verandas, or in boats on the beach. Amongst homeless users, heroin use takes place in other locations, such as a derelict house or sometimes on the streets in the open. Injecting at home avoids inconvenience and reinforces the message that one has a home. Using alone denotes that one has sufficient funds, and is therefore status enhancing, while not having one's own injecting equipment and borrowing or stealing from another user denotes a lack of control over one's life and a lack of autonomy [13].
Many users who inject in semi-public settings are embarrassed by their injecting practices, perhaps sharing the widespread general aversion to needles and syringes or perhaps, because according to their own local knowledge, injecting denotes a deeper level of dependency than smoking. This embarrassment or shame concerning injecting practices is far from unique [28]. On the other hand, some users, like Munira and Ibrahim, seek out opportunities to inject in public settings or walk around the neighbourhood with a syringe sticking out of their arms. This public display is not status enhancing because it denotes a lack of self-control. Sitting or standing in the semi-conscious state (kuyoyoma) in a public place, is a practice that is looked down upon by many users.
By 2003 only one pharmacy in Malindi would sell needles and syringes to heroin users. There have been a number of deaths of Malindi users, all injectors or ex-injectors. Nevertheless, it can be expected that the majority were sufferers of AIDS. Although in Malindi, the subject of HIV and AIDS remains 'taboo' amongst users and non-users alike; those who shared injecting equipment with now deceased members of their network have expressed their fear and despair to members of the Omari Project that they are also HIV positive.
Discussion
The permutations of sharing injecting equipment with single or multiple partners within different venues, such as the home, public spaces or a shooting gallery have implications for the spread of HIV [3,31]. In Europe and North America much work on HIV awareness has been carried out, and most IDU are aware that sharing injecting equipment is a very risky activity. However, sharing continues to occur, particularly between sexual partners and when a user is suffering from heroin withdrawal. Nevertheless, the easy availability of new needles and syringes through needles exchanges has done much to reduce levels of sharing equipment. The situation in Kenya is different.
There is limited awareness of the dangers of HIV infection from sharing injection equipment, and limited access to new or clean equipment. The widespread injection of heroin is a recent development, yet injecting practices have acquired social significance. This new culture of injecting heroin and the developing protocols of sharing have implications for the transmission of HIV. While high status individuals sometimes help out other users who are in withdrawal by providing a few free puffs of heroin, they would not routinely allow their equipment to be used or to share injected heroin with others. This emphasis on individual use, as opposed to sharing heroin and injecting equipment, is an aspect of mainstream Kenyan life where individual effort and enjoyment of the fruits of personal endeavour are the dominant survival strategy [33].
Heroin use in Coastal Kenya should be recognised as part of the process of economic and cultural globalisation. Although large scale trafficking in heroin between South Asia and East Africa is a recent phenomenon dating from the late 1970s, there is a long history of contact between nations bordering the Indian Ocean. In Kenya, heroin use is spreading to smaller towns and even remote villages. Hence, 'drug abuse' is reported to be 'rampant' in Mbajumwali [34], a small village on Pate Island within Lamu District on the North Kenya coast. This newspaper report confirms that heroin is easily available in Mbajumwali and the nearby bigger village of Kizingtini.
In other settings, the easy availability of new needles and syringes through needle exchanges has done much to reduce levels of sharing equipment, but this opportunity has not been available to injectors in Kenya. Studies from other parts of the world demonstrate that needle exchanges can reduce the transmission of HIV and other blood borne diseases [35-37], but that needle exchanges are most effective when part of 'an integrated set of diverse preventive measures' [38] as has been applied in Australia. Indeed, some studies in American and European cities demonstrate that needle and syringe exchanges have not prevented rises in HIV and Hepatitis B and C infections; such rises occurring among drug injectors in Vancouver [39], Montreal [40] and Amsterdam [41]. The reasons for the failure in prevention of HIV and Hepatitis B and C were the dilution of the impact of needle exchange schemes, due to an inadequate volume of syringes supplied (especially where cocaine injection was widespread as in Vancouver), insufficient focus on the risks of sharing injection paraphernalia, little impact on sexual behaviour and a non-interventionist ethos.
Developing needle exchange services in Kenya carries a clear responsibility to learn from such research. The findings of these studies indicate that the provision of adequate supplies of injecting equipment must be coupled with an interventionist ethos, whereby risk behaviour is reviewed regularly with injectors and strategies are developed to minimise these. This challenge is both exciting and immense.
Conclusion
Towards a Regional Response
Odek-Ogunde et al [5] found 44.9 % of heroin users in Nairobi had been or were injectors. In addition, 'white crest' is widely available in Nairobi and injection techniques are similar to those observed at the Coast (Odek-Ogunde, personal communication). Kenyan heroin users and development workers report that heroin is even more widely used and easily available in Nairobi and Tanzanian cities such as Zanzibar, Arusha and Dar es Salaam, than at the Kenyan Coast. The recent study from Dar es Salaam, Tanzania [7] indicates that injecting drug use is widespread in the Tanzanian capital. Similarly, the 2001 assessment of drug use in Tanzanian towns [6] found injecting drug use to be a matter of serious concern and advocated the setting up of needle and syringe exchange programmes. However, injecting drug use in Tanzania extends beyond the areas assessed in the 2001 situational analysis. For example, anecdotal reports by the Zanzibar Youth Forum (personal communication) indicate that rates of injecting are higher on the more remote island of Pemba than in Zanzibar town (the islands of Pemba and Unguja comprise Zanzibar, which is also the name of the main town on the island of Unguja).
Heroin users at the Kenya Coast, as elsewhere within the region, lack information about the dangers of injecting, the risks of sharing needles and syringes and unprotected sex. Odek-Ogunde has called for harm reduction measures and notes that 'the presence of IDUs in Kenya creates a three fold risk: the escalation of injection drug use, the potential of exacerbating the HIV/AIDS situation, and the creation of a source for other drug-related damage' [42].
Coastal Kenyan heroin injectors share injecting equipment and have sex with each other as well as with non-users. Their risky behaviour takes place in a setting where about 20% of the general population are estimated to be HIV positive [13]. The need for harm reduction initiatives amongst this high-risk group is clear. The findings of the study in Nairobi, which show that 1 in 2 injectors interviewed is HIV Positive, demand an urgent response [5]. Heroin users lack information about the dangers of injecting, of sharing needles and syringes, and of unprotected sex. Most East African towns have a lively sex industry and indeed the vast majority of women users work in this field.
Best practice from around the world demonstrates that HIV prevention through harm reduction measures should continue to target 'high risk groups' even when the epidemic has moved into the general population [2,4].
In November 2003, the Omari Project obtained funding from the UK Community Fund to open a drop-in centre in Malindi where harm reduction strategies, including needle exchanges will be pioneered. This offers an opportunity to curtail the transmission of HIV and other infections such as Hepatitis C, the prevalence of which is unknown in Kenya, although data from Nairobi indicates that both are very prevalent. The development of this intervention also offers the opportunity to learn from studies of needle exchange 'failures' in Europe and North America and to avoid repeating their unwitting mistakes.
The UK Department for International Development (DFID) has expressed interest in supporting such initiatives and extending them to other East African cities, and is providing funding to the Omari Project to assess injection drug use in Mombasa, with a view to developing appropriate harm reduction services, including a needle exchange. As in most parts of the world [29], functional needle exchange programmes will require careful advocacy work if the police and community are not to oppose their operation. The ESRC study in collaboration with the Omari Project, reported above, has already held a series of workshops with health workers and key community members in Malindi, where the dangers of injecting drug use and its relationship with HIV have been highlighted [14].
Acknowledgements
We thank the ESRC for funding this research (Grant no ROOO 23 8392). We also thank the men and women users who participated in the study. The Omari Project and Bristol Drugs Project thank the UK Community Fund for their continuing support. DFID and their managing agents Futures Group Europe are also thanked for their support in the pioneering of harm reduction approaches in East Africa.
==== Refs
Mwenesi HA Rapid assessment of Drug Use in Kenya Bulletin on Narcotics 1996 XLVIII 265 278
Ball AL Rana S Dehne KL HIV prevention amongst injecting drug users: responses in developing and transitional countries Public Health Report 1998 113 170 181
Des Jarlais DC Dehne K Casabona J HIV surveillance among injecting drug users AIDS 2001 15 S13 S22 11421178 10.1097/00002030-200104003-00003
Stimson G Choopanya K Stimson G, Des Jarlais DC Global Perspectives on Drug Injecting Drug Injecting and HIV Infection 1998 Ball A London: UCL Press
Odek-Ogunde M Lore W Owiti FR Risky behaviours among injecting drug users in Kenya Proceedings of 14th International Conference on the Reduction of Drug Related Harm 2003 Chiang Mai, Thailand
Kilonzo GP Mbwambo JK Kaaya SF Hogan NM Rapid Situational Assessment for Drug Demand Reduction in Tanzania UNDCP 2001
Muhondwa E Mpembeni R Mayunga F An assessment of the treatment needs of drug users un Dar es Salaam Report by Christ Compassion in Action prepared for Save the Children UK, Tanzania Programme 2002
Beckerleg S Telfer M Kibwana Sizi A Private struggles, public support: rehabilitating heroin users in Kenya Drugs: Education, Prevention and Policy 1996 3
Beckerleg S Counselling Kenyan Heroin Users: Cross Cultural Motivation? Health Education 2001 101
Beckerleg S Brown Sugar or Friday Prayers: Youth choices and community building in Coastal Kenya African Affairs 1995 94 23 38
UNDCP on The Eastern African Drug Beat Importance of Data in Control of Illicit Drug Trafficking 2000 2 13 14
Republic of Kenya on AIDS in Kenya 2001 Sixth Ministry of Health
Beckerleg S Lewando Hundt G (a) The characteristics and recent growth of heroin injecting in a Kenyan coastal town Addiction Research and Theory Forthcoming
Beckerleg S Lewando Hundt G L Hume, J Mulcock (b) Reflections on Fieldwork amongst Heroin Users in a Small town in Kenya Awkward Spaces, Productive Spaces: Fieldwork in Difficult Conditions Columbia University Press Forthcoming
Griffiths P Gossop M Powis B Strang J Reaching hidden populations of drug users by privileged access interviewers: methodological and practical issues Addiction 1993 88 1617 1626 8130701
Van Meter KM Lambert E Methodological and Design Issues: Techniques for Assessing the Representatives of Snowball Samples The Collection and Interpretation of data from hidden populations 1990 NIDA: Research Monograph Series 98: Rockville: NIDA
Power R A model for qualitative action research amongst illicit drug users Addiction Research 1995 3 165 181
Moore D Ethnography and Illicit Drug Use: dispatches from an anthropologist in the 'Field' Addiction Research 1993 1
Power R Rapid Assessment of the Drug-injecting situation at Hanoi and Ho Chi Minh City, Viet Nam Bulletin on Narcotics 1996 XLVIII 35 52 9839034
Power R Reflections on participant observation in drugs research Addiction Research and Theory 2001 9 325 337
Kaplan C Verbraeck H Where have all the fieldnotes gone? The changing nature and politics of drugs ethnography in The Netherlands Addiction Research and Theory 2001 9 229 323
Nuffield Council of Bioethics on The ethics of research related to healthcare in developing countries 2002
Simmons J Koester K Hidden injuries of research on social suffering among drug users Practicing Anthropology 2003 25 53 57
Wiebel W Lambert E Identifying and Gaining Access to Hidden Populations The collection and interpretation of data from hidden populations 1990 NIDA, Research Monograph Series 98: Rockville: NIDA
Harwood H Fountain D Carothers S Gerstein D Johnson R Stevens SJ, Wexler HK Gender differences in economic impacts of clients before, during and after substance abuse treatment Women and Substance Abuse 1998 London: The Haworth Press
Ramos R Lambert E Chicano Intravenous Drug Users The collection and interpretation of data from hidden populations 1990 NIDA, Research Monograph Series 98: Rockville: NIDA
Bourgois P Search of Respect: Selling Crack in El Barrio 1995 Cambridge: Cambridge University Press
Lockley P Counselling Heroin and Other Drug Users 1995 London: Free Association Books
Santelices C Singer M Nicolaysen AM Risky and precarious dependencies of Puerto Rican "IDUS" in El Barrio: an ethnographic glimpse Practicing Anthropology 2003 25 23 27
Beckerleg S Coomber R, South N Living with Heroin at the Kenya Coast Drugs in the Global Context: comparative perspectives on cultures and controls Free Association Press Forthcoming
Power R Jones S Kearns G Ward J An ethnography of risk management amongst illicit drug injectors and its implications for the development of community-based interventions Sociology of Health and Illness 1996 18 86 106 10.1111/1467-9566.ep10934419
Hughes RA Drug injectors and the cleaning of needles and syringes European Addiction Research 2000 6 20 30 10729739 10.1159/000019005
UNODCCP on The Drug Nexus in Africa 1999 UN Office for Drug Control & Crime Prevention Monographs, Vienna
Ndurya M Red Alert on Drug Deaths Coast Express 2003 1 7
Lurie P Drucker E An opportunity lost: HIV infections associated with lack of a needle-exchange program in the USA Lancet 349 604 608 1997, March 1 9057732 10.1016/S0140-6736(96)05439-6
Lurie P Reingold A School of Public Health on The public health impact of needle exchange programs in the United States and abroad 1993 Berkley, CA
Ksobiech A meta analysis of needle sharing, lending and borrowing behaviours of nep attenders AIDS: Education and Prevention 2003 15
Bastos F Stimson GV Telles P Barcellos C Stimson G, Des Jarlais DC, Ball A Cities responding to HIV-1 epidemics among injecting drug users Drug Injecting and HIV Infection 1998 London: UCL Press 149 167
Schechter MT Strathdee SA Cornelisse PJA Currie S Patrick DM Rekart ML O'Shaughnessy MV Do needle exchange programmes increase the spread of HIV among injection drug users? An investigation of the Vancouver outbreak AIDS 1999 13
Bruneaus J Langendam MW Notenboom J Coutinho RA High rates of HIV infection among injection drug users participating in needle exchange programmes in Montreal: results of a cohort study American Journal of Epidemiology 1997 146
Van Ameijden EJC Bruneau J Lamothe F Franco E Lachance N Desy M Soto J Vincelette J Continuing injecting risk behaviour: results from the Amsterdam cohort study of drug users Addiction 1999 94
Odek-Ogunde M Harm reduction as alternative strategy for prevention of drug abuse and HIV/AIDS in Kenya Proceedings of Third Annual Meeting of Global Research Network Meeting on HIV Prevention in Drug-Using populations 2000 Durban, South Africa 113 117
|
16143038
|
PMC1236950
|
CC BY
|
2021-01-04 16:38:14
|
no
|
Health Qual Life Outcomes. 2005 Sep 5; 3:52
|
latin-1
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-52
|
oa_comm
|
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-551614657310.1186/1477-7525-3-55ResearchResponse shift masks the treatment impact on patient reported outcomes (PROs): the example of individual quality of life in edentulous patients Ring Lena [email protected]öfer Stefan [email protected] Frank [email protected] David [email protected]'Boyle Ciaran A [email protected] Department of Psychology, Royal College of Surgeons in Ireland, Mercer Street Lower, Dublin 2, Ireland2 Dept of Pharmacy, BMC, Box 580, 751 23 Uppsala, Sweden3 Department of Medical Psychology and Psychotherapy, Medical University Innsbruck, Innsbruck, Austria4 Department of Restorative Dentistry & Periodontology, Dublin Dental Hospital, Dublin 2, Ireland5 The School of Dental Sciences, Trinity College, Dublin 2, Ireland2005 7 9 2005 3 55 55 11 3 2005 7 9 2005 Copyright © 2005 Ring et al; licensee BioMed Central Ltd.2005Ring 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
Quality of life (QoL) is now established as an important outcome for evaluating the impact of disease, and for assessing the efficacy of treatments. However, individuals change with time and the basis on which they make a QoL judgement may also change, a phenomenon increasingly referred to as response shift. Here, the individual may change his or her internal standards, values, and/or conceptualization on the target construct as a result of external factors such as a treatment or a change in health status. This has important implications for assessing the effects of treatments as a change in QoL may reflect a response shift, a treatment effect, or a complex combination of both. In this study, we used an individualised quality of life (IQoL) measure, the SEIQoL, together with a then-test to determine whether response shift would influence the measurement of treatment efficacy in edentulous patients.
Methods
Data are reported here for the first phase of a randomised controlled clinical trial designed to assess the impact, on IQoL, of implant supported dentures compared with high quality conventional dentures. IQoL was measured using the SEIQoL-DW in 117 patients (mean age 64.8; 32% male) at baseline (T1) and 3 months (T2) after receiving high quality conventional dentures. The work was carried out in dental teaching hospitals in Dublin and Belfast.
Results
Unadjusted SEIQoL index scores revealed no significant impact of treatment at three months (baseline: 75.0; 3 months: 73.2, p = .33, n.s.). However, the then-test at 3 months revealed that patients retrospectively rated their baseline IQoL as significantly lower (P < .001) than they had rated it at the time (then-test baseline: 69.2). Comparison of the 3 month scores with this readjusted baseline indicated a significant treatment effect (then-test baseline: 69.2; 3 months: 73.2, p = 0.016). 81% of patients nominated at least one different IQoL domain at 3 months.
Conclusion
The positive impact of denture treatment for edentulous patients on IQoL was seen only when response shifts were taken into consideration. The nature of the response shifts was highly complex but the data indicated a degree of re-conceptualisation and reprioritisation. Assessment of the impact of treatments using patient-generated reports must take account of the adaptive nature of patients.
==== Body
"I know who I was when I got up this morning, but I've changed several times since then." (Alice in Wonderland by Lewis Carroll)
Background
Patient reported outcomes (PROs) are widely used in clinical trials to incorporate patients' perspectives, and as adjuncts to "harder" measures of morbidity and mortality [1]. While PROs have long been considered simple, valid and reliable measures of outcome, it is increasingly clear that the cognitive processes involved in completing them are complex [2]. Recently attention has been drawn to potentially highly significant phenomena known as response shift [3]. As human beings, we actively construct meaning from our environment, and display a range of cognitive mechanisms to continually adapt to changing circumstances. Response shift refers to a change in the meaning of one's evaluation a construct as a result of a change in one's internal standards of measurement, a change in one's values, or a change in one's definition of the construct [3]. This means that persons might give different answers on PRO measures over time, not only because their health or quality of life (QoL) has changed, but also because they might have changed their perception on what health or QoL means to them. This may be particularly important in repeated measures trials where efficacy is measured as the change from a pre-treatment baseline following treatment.
QoL is highly individual with patients varying considerably in what they consider important for their QoL. It is also know that the elements of QoL change over time and in response to changing circumstances [4]. Individualised measures of QoL (IQoL) increase respondents' discretion in selecting the domains most important to their QoL and, in determining the relative importance of these domains. While it is feasible to use IQoL measures [4], they are not yet widely used in clinical trials [5]. IQoL measures such as the Patient Generated Index (PGI) [6] and the Schedule for Evaluation of Individual Quality of Life (SEIQoL) [7] may prove useful in determining the impact of response shift on the assessment of treatment effects in clinical trials.
The present study was designed to evaluate response shift in edentulous patients undergoing treatment. This is a chronic dental condition defined as total tooth loss and which can have a significant impact on QoL. Patients seek treatment for aesthetic reasons and to restore their oral function. Whereas biological outcome measures such as pocket depth, bone loss and chewing ability are important in this patient group, there is an increasing emphasis on the use of satisfaction and QoL measures to assess the impact of treatment [8]. Previous research has shown low correlations between patients' evaluations of their prostheses and clinicians' biological assessments [9].
Modern approaches to treatment involve replacing conventional dentures with dentures supported by osseointegrated implants set into the bone. Improvements in oral health following implants are well documented [10] and implant-supported dentures are considered superior to conventional dentures, since they are experienced as a part of the patient's own body and allow patients to feel food textures. Implants are more expensive than conventional dentures and the treatment is also somewhat more cumbersome since it involves dental surgery to attach the implants. Few longitudinal studies have used QoL as an outcome to compare implants and conventional dentures [8]. This study was designed as a randomised clinical trial comparing treatment with implants with high quality conventional dentures with QoL as the main outcome measure. Data are reported here only for the first phase of the study in which all patients wore new high quality (best possible) conventional dentures for three months before being randomised either to continue with the conventional dentures or to receive implants. The fitting of excellent dentures was done to ensure a proper baseline for the clinical trial since many edentulous patients report problems with ill-fitting dentures. This is the first longitudinal study to use the IQoL measure SEIQoL-DW [11] in this patient group.
We hypothesised that fitting high quality dentures or osseointegrated implants was likely to result in improved eating, communication, appearance and social life [12] and that consequently there should be a significant improvement in individual IQoL. Since the treatment periods were long (three months and six months), the psychological impact of treatment was likely to be marked on the main outcome measure individualised QoL. Therefore we decided to attempt to measure response shifts, since should these occur, they would complicate the interpretation of the IQoL data.
Methods
Data presented here were collected as part of a larger randomised controlled clinical trial comparing conventional dentures with osseointegrated implant supported dentures in edentulous patients. The Ethics committees of the respective dental teaching hospitals in Dublin and Belfast granted ethical approval. Eligible patients were recruited from waiting lists and from dental practices. Patients, identified by consecutive sampling, were invited to participate until the planned sample size of 70 patients per centre was reached. Patients had to have been edentulous for at least two years, be under 75 years of age, be medically suitable for surgery, be either non-smokers or smoking fewer than 10 cigarettes per day, have bone height at the anterior mandible on radiographic assessment of at least 1 cm, have the cognitive ability to understand the purpose of the study, and they had to provide informed consent. All patients were fitted with the highest possible quality conventional dentures at baseline and they wore these for three months before being randomly assigned either to continue with the conventional dentures or to receive implant supported dentures. This was designed to establish a clinically acceptable baseline treatment to assess the added value, if any, of implant supported dentures.
Individualised quality of life (IQoL)
IQoL was measured using the Schedule for the Evaluation of Individual Quality of Life – Direct Weighting (SEIQoL-DW) [11]. Respondents were first asked to nominate and describe the 5 areas of their lives (cues) that they consider to be the most important for their QOL. They were then asked to rate their current level of satisfaction/functioning on each cue on a scale between worst possible and best possible. Finally, they were requested to allocate 100 points to indicate the relative importance of each cue by using a pie-chart disc. The SEIQOL Index summary score, was derived by multiplying each cue's weight by its corresponding level, and summing the products across the 5 cues. The SEIQOL Index score ranged from 0 – 100, where a higher score indicates better QoL.
Response shift
The SEIQOL allows patients to nominate different cues at each assessment. It was assumed that, patients who nominated different cues as being important to their QoL at 3 months has changed their concept of what constituted QoL for them. This may reflect what Schwarz and Sprangers refer to as re-conceptualization (Table 1).
Table 1 SEIQoL-DW components of response shift
Response shift Description SEIQoL-DW indication
Re-conceptualisation Change in one's definition of the target construct Changes in nomination of cues (life areas) when comparing pre- and post-test cue nominations.
Re-calibration Change in one's internal standards of measurement Changes in cue (life area) levels when comparing the pre-test and then-test scores
Changes in SEIQoL Index scores when comparing pre-test and then-test scores
Re-prioritisation Change in one's values Change in cue (life area) weights when comparing the pre- and then-test scores
In addition to assessing re-conceptualization in the form of different cue nominations, we also wanted to determine whether other types of response shift might occur which could impact on the assessment of the treatment. Treatment effects are usually determined by assessing changes in scores from a pre-treatment baseline. It is assumed that a change represents a treatment effect. However, if a respondent shows a response shift by changing the manner in which he or she completes the measure, this may confound interpretation of the treatment effect. The then-test has been proposed as a method that allows some aspects of this process to be assessed [13]. The patient is asked at T2 retrospectively to rate their situation at T1, not as they recall it, but as they now see it. The theory is that they should be using the same internal criteria for rating T1 as they are now using for rating T2. Differences between the original T1 rating and the retrospective T1 rating indicate response shift. If such differences are found, the difference between the retrospective T1 score and the T2 score is a more accurate indication of the treatment effect since it controls for the response shift. For example, a patient rates her pre-treatment level of pain as 7 on a 10-point pain scale. She subsequently rates her post-treatment level of pain as 3. This is taken to indicate that the treatment has caused an improvement of 4 points. However, if she retrospectively rates her pre-treatment pain as having been a 5, the actual treatment effect is 2. Likewise, if she retrospectively rates her pre-treatment pain as having been 10, the actual treatment effect is 7. The theory proposes that she is using the same internal standard to make the retrospective assessment as she is using to make the current assessment. It can be seen from this example that a response shift in a PRO may cause one to overestimate or underestimate the real effect of a treatment if one is basing the judgement solely on changes in raw scores.
We used the then-test at three months by asking patients to rate retrospectively their original cues at baseline. They were asked to re-rate cue levels (re-calibration) and cue weights (re-prioritisation). The hypothesised components of response shift measured in this way are shown in Table 1. The then-test was administered using the following wording:
"You have shown me how your quality of life is at the moment. The five important life areas that you have spoken about today are the same five areas that you chose when we first met1. I would like you to show me how you now think you were doing in each of these five life areas when we first met, by using this same scale that we used earlier. ...I am not asking you to try and remember how these important life areas were functioning, but rather how, when looking back today, you now think they were functioning when we first met."
Where patients had selected different cues at T2 they were asked in the then-test to retrospectively rate the cues originally chosen at T1.
The following wording were used to assess re-prioritisation:
"Now I would like you to show me how important you now think your five life areas were in relation to each other when we first met. Once again, I am not asking you to try and remember how important these life areas were in relation to each other, but rather what, when looking back today, you now think was their relative importance when we first met."
Three experienced dental nurses, trained in the SEIQoL interview technique and using a standard protocol for administration of the then-test, administered all assessments.
Statistics
Data were described using frequency distributions. Paired t-tests were used to assess changes from T1 to T2. Differences at a specific time point between variables were assessed by independent t-tests. P values of p < .05 were considered significant.
Results
Complete data were available for 117 patients (83.6%) at baseline (T1) and 3 months after (T2) receiving high quality conventional dentures. The mean age of the sample was 64 ± 8 years and 32% were male. The aim was to include 140 patients but the final sample consisted of 136 patients. Two patients died before follow-up, others dropped out due to having cancer (n = 1), or due to having sick or dying relatives (n = 2), or were withdrawn due to denture/implant problems (n = 4) and some dropped out without reason (n = 5). Of the 122 patients finally included at T1, 117 completed all study measures at both time points. None of the patients found it difficult to complete SEIQoL.
Cues nominated
The most frequently (percent of patients, choosing the category as cue 1 and 5) nominated cues were; Family/next of kin (46%–4%) e.g., partner, brother, mother, children, grandchildren; Health (34%–5%) e.g., stay healthy, being fit, being alive, pain; Hobbies/recreation (2%–31%) e.g., sports, reading, dancing, golf; Social life (2%–8%) e.g., meeting up with friends, ill fitting dentures preventing social life, embarrassed to eat and talk; and Religion/faith (3%–11%) e.g., God is important in life, like to believe there's something after life. Cues directly concerned with dental function were less frequently nominated (5%–3%) e.g., oral health, eating out, dentures, being able to eat/talk properly and appearance.
Treatment effect
As shown in Figure 1, there was no significant change in unadjusted SEIQoL-DW Index scores from baseline to 3 months following treatment (pre-test: 75.0; post-test: 73.2, p = .33). Baseline Index scores generated from retrospective assessments using the then-test were significantly lower that the Index scores generated using the original baseline data (original baseline: 75.0; then-test baseline: 69.2, p < .001). When Index scores at 3 months were compared with the baseline scores generated using the then-test, a significant improvement in IQoL following denture treatment was found (then-test: 69.2; post-test: 73.2, p = 0.016).
Figure 1 SEIQoL Index scores at baseline (T1) and after three months (T2). 1The traditional reported treatment effect = post-test minus pre-test score 2The response shift effect = pre-test minus then-test score 3The actual treatment effect = post-test minus then-test score
Response shifts
Cues
81% of patients nominated at least one different cue at 3 months compared to baseline.
Cue levels
In the then-test, all patients were asked to retrospectively rate their baseline cue levels and weights at 3 months even if they had nominated new cues at the latter time. As shown in Table 2, mean cue levels ranged from 76-73 (pre-test), 68–75 (post-test) and 68–70 (then-test) on the scale from 0–100. In the then-test, patients significantly changed their ratings of the cue levels of 4 of 5 cues, on average, indicating a significant degree of re-calibration. There was a significant change from baseline at 3 months in level of the most important cue (the cue given the highest relative weight), when comparing post-test and then-test scores. While the most important cue nominated at 3 months is not necessarily the same as that nominated at baseline, the change in the average level of the most important cue may reflect a significant treatment effect.
Table 2 Average cue levels and weights for each of the 5 SEIQOL-DW cues selected by respondents at baseline.
Cues* T1 – pre-test mean level ± SD.
mean weight ± SD. T2 – post test mean level ± SD.
mean weight ± SD. T1 – then-test mean level ± SD.
mean weight ± SD. P-value Levels # P-value Weights #
1 pre-post: .359 pre-post: .026
level 73.8 ± 28.3 74.6 ± 22.3 69.7 ± 26.9 pre-then: .076 pre-then: .033
weight 29.6 ± 12.4 26.9 ± 9.2 26.7 ± 9.79 post-then: .005 post-then: .649
2 pre-post: .088 pre-post: .404
level 76.2 ± 24.1 71.1 ± 26.2 69.0 ± 24.7 pre-then : .003 pre-then .152
weight 23.3 ± 10.59 22.3 ± 9.47 21.8 ± 8.61 post-then: .106 post-then: .598
3 pre-post: .069 pre-post: .801
level 75.7 ± 27.1 70.8 ± 24.0 70.1 ± 24.7 pre-then: .006 pre-then: .534
weight 20.2 ± 9.91 20.6 ± 8.85 21.0 ± 8.97 post-then: .403 post-then: .808
4 pre-post: .269 pre-post: .223
level 72.7 ± 24.7 68.1 ± 26.4 67.9 ± 23.8 pre-then: .036 pre-then: .073
weight 15.0 ± 7.91 16.2 ± 8.27 16.6 ± 7.75 post-then:.424 post-then:.576
5 pre- post:.105 pre- post:.012
level 74.2 ± 22.6 69.5 ± 25.7 68.7 ± 23.9 pre-then: .028 pre-then:.015
weight 12.3 ± 6.05 14.2 ± 7.03 14.3 ± 6.50 post-then: .635 post-then: .777
* Cues in descending order of importance at T1
# The traditional reported treatment effect = post-test minus pre-test score
# The response shift effect = pre-test minus then-test score
# The actual treatment effect = post-test minus then-test score
Cue weights
Mean cue weights (see Table 2) ranged from 12%–30% (pre-test) and 14%–27% (post-test) and 14%–27% (then-test). In the then-test (comparing pre-test and then-test scores), patients significantly changed their weightings, on average, for the most and least important cues. At 3 months, the weights assigned to the most and least important cues were significantly different than those assigned to the most and least important cues at baseline (Note: the most important and least important cues nominated at 3 months are not necessarily the same as those nominated at baseline).
Discussion
The main focus of this paper is the significance of response shifts for assessing treatment outcomes. Disparities between objective clinical measures and patients' subjective assessments are common. Patients with the same condition respond differently and even the same patient can respond differently over time. QoL measures used currently in clinical research were not designed to account for response shifts but are based on the assumption that people respond consistently on measurement scales and also that scales are directly comparable across individuals and over time. The classical approach has been to consider individual differences in response as sources of error. However, Schwartz and Rapkin [14] have argued convincingly that individual differences in cognitive appraisal processes should be viewed, not as sources of error in QoL research but, rather, that these properties are intrinsic to all QoL measurement.
In this study, we used an individualised measure of quality of life, the SEIQoL-DW, as we felt that, by focusing on the unique choices of patients, we would be in a position to detect more clearly any response shifts that might occur. SEIQoL index scores did not reveal a significant improvement in IQoL 3 months after receiving high quality conventional dentures. However, when the baseline scores were derived based on the then-test, and when comparing then-test and post-test estimates, a significant improvement was seen. Response shifts had occurred in that patients had changed their criteria for assessing their quality of life from baseline to 3 months. It was only when this change was factored into the analysis that the improvement following treatment could be seen. The changes in the SEIQoL were highly complex but it is possible to gain some insight into their nature by looking at the various components of the measure i.e. cues, weights and levels.
Four out of every 5 patients (81%) nominated at least one different QoL cue at 3 months compared to baseline. Therefore, the elements that they considered most important for their quality of life changed over the study period. This represents a form of re-conceptualisation, one with which clinicians will be familiar. Patients change and adapt with time and in response to changing circumstances. The domains that might have been important for one's QoL before treatment may not be as important on a subsequent occasion. The same phenomenon can be seen with disease progression. Some patients with severe chronic conditions report higher QoL than do healthy individuals [15]. Significantly disabled or terminally ill patients sometimes report QoL similar to or higher than that of healthy controls [16]. One limitation of the SEIQoL-DW in this context is that, the respondent is only allowed to select 5 cues. If she chooses different cues on a subsequent occasion from those chosen previously, it could be argued (as we have done) that she has re-conceptualised what QoL means to her. But if she were allowed select as many cues as she wished and she included all of the cues previously chosen as well as any new ones, then this would be more likely to indicate re-prioritisation rather than re-conceptualization. Patients may also have used different words at each evaluation to refer to essentially the same area. This can be controlled by collecting detailed descriptions of the life areas chosen as well as including questions assessing patients' own perception of change.
Patients were asked at 3 months to indicate retrospectively their level of functioning on each of the cues chosen at baseline. In general, patients retrospectively rated their level of functioning on most of the cues as lower that they had done at the time. If we assume that they completed both assessments at 3 months using a single internal frame of reference, it seems reasonable to label this as re-calibration. It may be that the superior function associated with the quality dentures provided caused patients retrospectively to perceive their pre-treatment levels as worse on reflection.
Because the SEIQoL-DW weights are individualised, it is possible to measure changes in the relative importance of cues over time. We found that on average some weights (comparing pre-test and then-test weights) changed indicating that reprioritisation can occur. However, when comparing then-test and post-test weights we found no changes. This might be a true finding, or maybe patients simply applied the same weights they were using at T2 to the cues at T1. This may also partly be an artefact of the SEIQoL-DW procedure as the weights of all five cues selected by respondents are constrained to add to 100. Therefore, if the relative importance of one cue increases, the relative importance of at least one of the other four cues must diminish.
One of the major challenges in interpreting the results of this study is that 81% of the patients chose at least one different cue at 3 months compared to baseline. All patients were asked, in the then-test, to re-evaluate their baseline cues, whether they were the same or not. It seems likely that this process is different for those who chose the same and different cues at 3 months and this is worthy of further research. The sample size of 19% of patients who chose exactly the same cues at 3 months was too small to draw firm conclusions about the nature of response shift in this group.
Some studies have found that memory can influence the findings from the then-test [17]. A limitation with our study is that we did not control for recall bias and we did not compare the changes with any criterion measure of change [18,19]. However, receiving dentures is a significant and salient event and it seems likely that the influence of recall bias is minimal especially given the number of judgements a patient had to make and the 3 month gap between assessments. One alternative explanations for our findings of a discrepancy between prospective and retrospective assessments is that subjects may have expected that receiving high quality dentures should improve their health, an they retrospectively rated their initial health as lower to reflect this expectation, a cognitive mechanism known as the implicit theory of change [20]. Our interpretation of the results is based on the assumption that the retrospective then-test data provides a more valid indication of baseline IQoL for comparison with 3 month data than does the baseline assessment itself. If, however we assume that the retrospective judgement is biased and that the concurrent baseline assessment is more valid, our results would be interpreted differently and there would be no treatment effect. To support the response shift theory, we would need to show that the new information available to patients after receiving their dentures led to more valid judgments of their baseline scores. However, it is as yet unclear how one would determine which theory is more valid for a particular situation. It would be important to distinguish patients who's situation had improved or deteriorated from those who had changed their mind about what it means to have the best or worst possible outcome.
Recently, Schwartz and Rapkin have proposed a new psychometric model, which posits that the "true" PRO score is contingent on aspects of the appraisal process [14]. The appraisal of a construct like QoL may be related to culture, personality and situation and may vary across persons and over time [21-24]. Building on the response shift model, Schwartz and Rapkin have proposed using an Appraisal Profile [21]. They suggest that "rather than simply asking people to re-rate their baseline status using "today's criteria", we assess their appraisal processes to make those criteria explicit at each time in order to help characterise qualitative change". Improved knowledge about the ways in which patients appraise QoL might lead to more valid, reliable and responsive measures. Future studies need to disentangle the differing ways individuals appraise QoL and researchers must acknowledge the dynamic nature of QoL by empirically testing for response shift phenomena.
Conclusion
Improvements in the IQoL in edentulous patients, following treatment with high quality dentures, were apparent only when patient adaptation over time, was taken into account. This study demonstrated that an IQoL measure such as the SEIQoL-DW can be used to assess re-conceptualisation and reprioritisation and can be applied as a then-test to control for recalibration.
List of abbreviations
QoL quality of life
IQoL individualised quality of life
SEIQoL-DW Schedule for the Evaluation of Individualised Quality of Life – Direct Weighting
PRO patient reported outcomes
Authors' contributions
CAO, DH and FH developed the core idea. CAO, DH and FH designed the study. LR and SH monitored the study. CAO, LR and SH conducted the literature search. SH and LR performed the statistical analyses. LR wrote the first draft of the paper. All authors critically reviewed and contributed to the final draft of the paper and all are guarantors.
Acknowledgements
This study was funded by Straumann Ltd, Switzerland. The founding source had no involvement in either study design or in the collection, analysis and interpretation of data or in the writing of the report and the decision to submit the paper for publication.
This manuscript was completed while the authors Dr. Lena Ring and Dr. Stefan Höfer were EU Marie Curie Research Fellows at the Department of Psychology, Royal College of Surgeons in Ireland.
We also wish to acknowledge the Royal Victoria Hospital, School of Dentistry, Belfast for patient recruitment and dental nurses Rosaleen Glackin, Sharon Kennedy and Janine Burns for performing the SEIQoL interviews. Finally, a special acknowledgement to Sarah Clarke for assistance with initial planning and setting up of the study.
==== Refs
Wiklund I Assessment of patient-reported outcomes in clinical trials: the example of health-related quality of life Fundam Clin Pharmacol 2004 18 351 363 15147288 10.1111/j.1472-8206.2004.00234.x
McColl E Meadows K Barofsky I Cognitive aspects of survey methodology and quality of life assessment Qual Life Res 2003 12 217 218 12769133 10.1023/A:1023233432721
Sprangers MA Schwartz CE Integrating response shift into health-related quality of life research: a theoretical model Soc Sci Med 1999 48 1507 1515 10400253 10.1016/S0277-9536(99)00045-3
Joyce CRB McGee HM O'Boyle CA Individual quality of life: Approaches to conceptualisation and assessment 1999 Amsterdam, Harwood Academic Publishers
O'Boyle C Höfer S Ring L Fayers P and Hays RD Individualized Quality of Life Assessing the quality of life in clinical trials 2004 2 Oxford Oxford University Press 225 242
Ruta DA Garratt AM Leng M Russell IT MacDonald LM A new approach to the measurement of quality of life. The Patient-Generated Index Med Care 1994 32 1109 1126 7967852
O'Boyle CA McGee H Hickey A O'Malley K Joyce CR Individual quality of life in patients undergoing hip replacement Lancet 1992 339 1088 1091 1349111 10.1016/0140-6736(92)90673-Q
Allen P Assessment of oral health related quality of life Health Qual Life Outcomes 2003 1 40 8 September 2003 14514355 10.1186/1477-7525-1-40
Feine JS Awad MA Lund JP The impact of patient preference on the design and interpretation of clinical trials Community Dent Oral Epidemiol 1998 26 1 5
Boerrigter E Stegenga B Raghoebar G Boering G Patient satisfaction and chewing ability with implant retained mandibular overdentures Int J Oral Maxillofacial Surg 1996 53 1167 1173 10.1016/0278-2391(95)90627-4
Hickey AM Bury G O'Boyle CA Bradley F O'Kelly FD Shannon W A new short form individual quality of life measure (SEIQoL-DW): application in a cohort of individuals with HIV/AIDS BMJ 1996 313 29 33 8664768
Allen PF McMillan AS A review of the functional and psychosocial outcomes of edentulousness treated with complete replacement dentures J Can Dent Assoc 2003 69 662 14611716
Sprangers MAG van Dam FSAM Broersen J Revealing response shift in longitudinal research on fatigue: The use of the thentest approach Acta Oncol 1999 38 709 718 10522761 10.1080/028418699432860
Schwartz CE Rapkin BD Reconsidering the psychometrics of quality of life assessment in light of response shift and appraisal Health and Quality of Life Outcome 2004 2 16 23(March2004) 10.1186/1477-7525-2-16
Ubel PA Loewenstein G Jepson C Whose quality of life? A commentary exploring discrepancies between health state evaluations of patients and the general public Qual Life Res 2003 12 599 607 14516169 10.1023/A:1025119931010
Bach JR Tilton MC Life satisfaction and well-being measures in ventilator assisted individuals with traumatic tetraplegia Arch Phys Med Rehabil 1994 75 626 632 8002759 10.1016/0003-9993(94)90183-X
Ahmed S Mayo NE Wood-Dauphinee S Hanley JA Cohen SR Response shift influenced estimates of change in health-related quality of life poststroke J Clin Epidemiol 2004 57 561 570 15246124 10.1016/j.jclinepi.2003.11.003
Schwartz CE Sprangers MA Methodological approaches for assessing response shift in longitudinal health-related quality-of-life research Soc Sci Med 1999 48 1531 1548 10400255 10.1016/S0277-9536(99)00047-7
Schwartz CE Sprangers MAG Carey A Reed G Exploring response shift in longitudinal data Psychol Health 2004 19 51 69 10.1080/0887044031000118456
Norman G Hi! How are you? Response shift, implicit theories and differing epistemologies Qual Life Res 2003 12 239 249 12769136 10.1023/A:1023211129926
Rapkin BD Schwartz CE Toward a theoretical model of quality-of-life appraisal: Implications of findings from studies of response shift Health and Quality of Life Outcome 2004 2 14 15 March 2004 10.1186/1477-7525-2-14
Nisbett RE Peng K Choi I Norenzayan A Culture and systems of thought: holistic versus analytic cognition Psycol Rev 2001 108 291 310 10.1037//0033-295X.108.2.291
Diener E Oishi S Lucas RE Personality, culture, and subjective well-being: emotional and cognitive evaluations of life Annu Rev Psychol 2003 54 403 425 12172000 10.1146/annurev.psych.54.101601.145056
Suh E Diener E Oishi S Triandis HC The shifting base of life satisfaction judgements across cultures: emotions versus norms J Pers Soc Psychol 1998 74 482 493 10.1037//0022-3514.74.2.482
|
16146573
|
PMC1236951
|
CC BY
|
2021-01-04 16:38:14
|
no
|
Health Qual Life Outcomes. 2005 Sep 7; 3:55
|
utf-8
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-55
|
oa_comm
|
==== Front
Int J Behav Nutr Phys ActThe International Journal of Behavioral Nutrition and Physical Activity1479-5868BioMed Central London 1479-5868-2-111613893310.1186/1479-5868-2-11ReviewPerceived environment and physical activity: a meta-analysis of selected environmental characteristics Duncan Mitch J [email protected] John C [email protected] W Kerry [email protected] School of Health & Human Performance, Central Queensland University, Rockhampton, Australia2 Faculty of Physical Education and Recreation, University of Alberta, Edmonton, Canada2005 5 9 2005 2 11 11 21 2 2005 5 9 2005 Copyright © 2005 Duncan et al; licensee BioMed Central Ltd.2005Duncan 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
Several narrative reviews have been conducted on the literature examining environmental correlates of physical activity (PA). To date these reviews have been unable to provide definitive summaries of observed associations. This study utilizes meta-analytical techniques to calculate summaries of associations between selected environmental characteristics and PA.
Methods
Published studies were identified from electronic databases and searches of personal files. Studies were examined to determine the environmental constructs most frequently studied. Included studies (N = 16) examined at least one identified construct and determined associations between perceived environmental constructs and PA using logistic regression. Data were analyzed separately for crude and adjusted ORs using general-variance based fixed effect models.
Results
No significant associations emerged between environmental characteristics and PA using crude OR. The perceived presence of PA facilities (OR 1.20, 95% 1.06–1.34), sidewalks (OR 1.23, 95% 1.13–1.32), shops and services (OR 1.30, 95% 1.14–1.46) and perceiving traffic not to be a problem (OR 1.22, 95% 1.08–1.37) were positively associated with activity using adjusted ORs. Variance in PA accounted for by significant associations ranged from 4% (heavy traffic not a problem) to 7% (presence of shops and services).
Conclusion
Results of the meta-analysis support the relevance of perceived environmental characteristics for understanding population PA. These results should encourage the use of comprehensive ecological models that incorporate variables beyond basic demographic information.
==== Body
Introduction
The burden of disease attributable to physical inactivity is estimated at $377 million in Australia [1], $2.1 billion in Canada [2] and $24 billion in the U.S. [3], and continues to rise as large majorities of populations remain insufficiently active for health benefits. Research relating to the determinants of physical activity and inactivity has previously focused on determinants at the individual level, largely neglecting physical environments as influences of PA [4]. It is now acknowledged that environments that people build and inhabit provide potential opportunities and barriers to engaging in physically active lifestyles [5,6]. Furthermore, suburban sprawl and the way neighborhoods are designed are related to the physical health and bodyweight status of adults residing in those neighborhoods [7,8]. Thus, research understanding how elements of the natural and built environment influence PA is increasing [9-11].
Findings of existing primary studies and narrative reviews studying the associations between the perceived environment and PA are ambiguous. For example, pathways in close proximity to the home have been found to have a positive association with PA in some instances [12] while showing no association with PA in others [13]. Contrary to intuitive belief, positive associations have been found between the presence of unattended dogs and PA using both subjective [14] and objective [15] measures. Perceptions of neighborhood crime have been found to have both a negative association [15,16] and a lack of association [14] with PA. Similarly, the self-reported presence of street lighting has produced both positive [17] and no association [12,14] with activity using self-report measures and objectively defined geographic information system (GIS) measures [15]. Recent narrative reviews [11,18,19], while providing a useful summary of existing research, have been unable to resolve the aforementioned ambiguities.
Meta-analyses permit the strength and direction of associations between dependent and independent variables to be detailed [20,21]. In comparison to narrative reviews, meta-analyses have predefined inclusion criteria, consider sample size of the study when combining effects, and provide summaries of effects across different outcome measures [22]. In addition, narrative reviews may be subject to inclusion bias, as only studies supporting the reviewer's hypothesis may be included [23-25]. However, meta-analysis is not without its critics. For instance, it has been argued that combining effects from observational studies accentuates the methodological and statistical inaccuracies of included studies [26]. However, the application of appropriate statistical techniques and well defined inclusion criteria can help alleviate some of these concerns [27]. The purpose of this meta-analytic review is to identify the strength and direction of relationships between characteristics of the perceived environment and PA.
Methods
Search Procedures
Studies were located from five sources. First, searches of the electronic databases MEDLINE, Proquest, and Infotrac were conducted for the period from 1989 to February 2005. Search terms included, but not limited to, physical activity, exercise, walking, environment, built environment, and perceived environment. Second, a manual search was performed, for the same time period, on the following Journal titles: Medicine and Science in Sports and Exercise, Preventive Medicine, American Journal of Public Health, American Journal of Health Promotion, and American Journal of Preventive Medicine. These titles were included for manual searches as the majority of studies identified from electronic searches were published in these titles. Third, to identify previously unidentified studies, reference lists from narrative reviews on the topic [11,18,19] were examined. Fourth, a search of each author's personal files was conducted. Finally, a request was sent to an electronic PA list-serve (University of South Carolina Prevention Research Center – USCPRC) soliciting any articles accepted for publication and currently in press.
Study Inclusion Criteria
Studies were initially included in the meta-analysis if they were published in English and assessed any characteristic of the physical environment in relation to PA. Preliminary searches and coding revealed 50 published studies on the topic assessing 138 separate objectively and subjectively-measured environmental attributes. Each attribute was assessed in a minimum of 1 and a maximum of 15 studies. Because the majority of studies used measures of the perceived environment, the current review included only those constructs assessed using measures of the perceived environment. Finally, studies went through a three-stage inclusion process: 1) Environmental constructs represented by a minimum of at least five effect sizes (ES) were included in the meta-analysis (i.e., presence of PA facilities; presence of sidewalks, shops, and services; high levels of crime; street lighting; unattended dogs in the neighborhood); 2) Studies were included only if they assessed the association between the perceived environment and PA using logistic regression and reported sufficient details for ES calculation (i.e., Odds Ratios [OR] and 95% Confidence Intervals [CI]); and, 3) in cases where a series of papers reported on populations that were not independent of one another [14,28] only a single report that satisfied all inclusion criteria from the series was included in the analysis [14].
Data Extraction
All data were extracted independently by the first author. Where associations between characteristics of the physical environment and PA were reported by separate ethnic groups [14,29], or gender [30-32], the ORs were entered as separate cases. Included environmental constructs were coded separately within the analysis. In addition to OR and 95% CI for each case, several variables were coded for descriptive purposes, including characteristics of the sample (sample size, gender, ethnicity, level of sample [local, state, national], geographic location of sample [urban, rural, mixed], and methodology [theoretical basis, PA measure, mode of PA, reliability and validity of PA and environment measures]). The frequency of these descriptive characteristics are summarized by environmental characteristic in Table 1. Study Design (cross-sectional, experimental, longitudinal) were also coded, however, all studies satisfying inclusion criteria were cross-sectional. When studies reported variables as tertiles, the middle tertile was not included in the analysis in order to conserve the dichotomous nature of the environmental constructs [33-35]. In some instances the referent category used in the initial analyses was opposite to that used in the majority of other cases for the construct (i.e. "Yes" reference category when majority of studies used "No"). When this occurred (n = 5), such cases were removed from the calculation of the ES. While it is possible to reverse categories, the original fourfold table is required to recalculate the OR with the reversed referent category [36]. However these original fourfold tables were not reported in any of the excluded manuscripts.
Table 1 Contribution of the Coded Variable to Total Number of ESs Summarized Environmental Characteristics
Environmental Characteristic
Coded Variable PA Facilities Sidewalks Shops & Services in Walking Dist. Heavy Traffic is Present High Crime is Present Street Lighting is Present Unattended Dogs are a Problem Total
Gender
Men 6.67 - 9.09 6.25 - - - 3.03
Women 66.67 68.75 81.82 68.75 69.23 86.67 69.23 72.73
Mixed 26.67 31.25 9.09 25.00 30.77 13.33 30.77 24.24
Ethnicity
Combined 40.00 31.25 27.27 37.50 30.77 13.33 30.77 30.30
Nat. Amer. 6.67 12.50 9.09 6.25 7.69 13.33 7.69 9.09
Caucasian 6.67 12.50 9.09 6.25 7.69 13.33 7.69 9.09
African Amer. 26.67 25.00 27.27 25.00 30.77 33.33 30.77 28.28
Latino 20.00 18.75 27.27 25.00 23.08 26.67 23.08 23.23
Country of Sample
US 86.67 93.75 90.91 87.50 100.00 100.00 100.00 93.94
Australia - 6.25 9.09 12.50 - - - 4.04
England 13.33 - - - - - - 2.02
Geographic Location of Sample
Urban 26.67 25.00 81.82 25.00 30.77 26.67 30.77 33.33
Rural 13.33 12.50 18.18 12.50 15.38 13.33 15.38 14.14
Mixed 60.00 62.50 - 62.50 53.38 60.00 53.85 52.53
Theoretical Basis of Study
Ecological 60.00 75.00 54.55 75.00 69.23 86.67 69.23 70.71
SCT - - 18.18 - - - - 2.02
NS 40.00 25.00 27.27 25.00 30.77 13.33 30.77 27.27
PA Mode Assessed
Walking 13.33 - 18.18 - - - - 4.04
Trans. Walking - 6.25 9.09 - - - - 2.02
Sufficient LTPA 13.33 37.50 - 12.50 15.38 26.67 15.38 18.18
Sufficient PA 73.33 56.25 72.73 87.50 84.62 73.33 84.62 75.76
PA Measure
BRFSS 13.33 12.50 - 13.33 15.38 - 15.38 10.20
BRFSS + NPA 13.33 37.50 - 13.33 15.38 40.00 15.38 20.41
SEID - 6.25 9.09 - - - - 2.04
S-IPAQ - - - 13.33 - - - 2.04
7D-PAR - - 18.18 - - - - 2.04
NS 13.33 - - - - - - 2.04
WPA 60.00 43.75 72.73 60.00 69.23 60.00 69.23 61.22
Notes: Consult studies included in current analysis for citations of PA Measures used.
Percentages are reported as the percentage of effect sizes derived from coded variables in each environmental category.
NS – Not Specified
SEID – Study on Environmental and Individual Determinants of Physical Activity Measure
WPA – Women and Physical Activity Measure; NHI – National Health Interview
S-IPAQ – Short Form International Physical Activity Questionnaire
Because several studies report associations between environmental variables and PA using adjusted ORs, separate analyses were conducted for unadjusted and adjusted ORs. In the case of adjusted ORs, we coded for the variables that were adjusted for in the logistic regression.
Statistical Analysis
To stabilize the variance of OR and to adjust for the nature of OR (where values of 0.5 and 2 hold equivalent strengths in opposite directions), ORs were transformed to their natural logarithms (ln). The resultant lnOR are approximately normally distributed with positive and negative values representing positive and negative associations respectively [37]. Variance surrounding each OR were calculated using the following formula
Variance ORi = [ln(ORu/ORi)/1.96]2
Where ORu is the upper confidence interval of the ith OR, and ORi is the OR of the ith study. Confidence intervals and the crude OR from each case were used to estimate the overall OR for each environment variable using a general variance-based fixed effect model [38].
ORs = e∑[(wi × ln ORi)]/∑wi
Where ORs is the summary OR for the specified variable, wi is the weight of the ith study, which is the inverse of ith study's variance and ln ORi is the natural log of the OR for the ith study [38]. Homogeneity of variance was tested across studies by the Mantel-Haenszel method using the following formula:
Q = ∑[wi × (ln ORmh - ln ORi)2]
Where Wi is the weight of the ith study, and ORmh is the natural log of the summary OR for the dependent variable. The Q statistic provides a test of whether the group of effect sizes represents a common population effect. A significant Q statistic indicates the effect does not represent the population of studies and that a search for moderator variables is warranted. Variance in PA explained by each variable (R2) was calculated using the procedure proposed by Nagelkerke [39].
Results
Descriptive Summary of Sample Studies
Table 1 presents a summary of the coded characteristics that may influence relationships between perceptions of the environment and PA. Across all environmental categories, most of the included studies were conducted in the USA, two studies were conducted in Australia and one study was conducted in the U.K. Native Americans (n = 9) were the least studied ethnic population within the sample, and African Americans (n = 28) were the most studied ethnic population. Most effect sizes (n = 72) were drawn from female populations whereas few effect sizes (n = 3) were drawn from exclusively male populations. The majority of study populations were drawn from local geographic regions. Most studies utilized ecological models of health behavior; however, two studies [40,31] did not specify an underlying theoretical basis. Physical activity outcome variables were comprised largely of people meeting sufficient physical activity criteria [41], including sufficient leisure-time physical activity (n = 18) [14] and sufficient physical activity (n = 75) [42]. The prominent PA assessment method was the Women's Physical Activity Survey (n = 60) used in the Women's Cardiovascular Health Network Project [43]. This was the prominent PA assessment method as a large number of included studies were part of the Women's Cardiovascular Health Network Project.
Unadjusted Associations (Crude Odds Ratios)
Using crude ORs, no variables demonstrated a significant association with PA (Table 2).
PA facilities was the only variable to exhibit significant heterogeneity among effect sizes (Q[6] = 64.41, p < 0.05). Unfortunately, the limited number of effect sizes within PA facilities precludes any search for moderators to resolve this heterogeneity.
Adjusted Associations
Age, income and education level were the most commonly adjusted for variables in the original studies. When individually adjusted ORs were combined, four significant associations emerged. People who reported the presence of PA facilities were more likely to engage in PA than those not reporting proximal infrastructure (OR = 1.20, 95% CI. 1.06–1.34). Similarly, people reporting the presence of sidewalks, compared to those reporting the absence of sidewalks, were more likely to be physically active (OR = 1.29, 95% CI. 1.17–1.41). Those people reporting the presence of shops and services within the neighborhood were more likely to engage in activity compared to people not reporting their presence (OR = 1.30, 95% CI. 1.14–1.46). People reporting that heavy traffic was not a problem were more likely to engage in PA compared to those reporting heavy traffic was a problem (OR = 1.22, 95% CI 1.08–1.37). Examination of Q statistics revealed that PA facilities Q(7) = 22.58, p < 0.05, and the presence of shops and services Q(6) = 25.22, p < 0.05, both displayed significant heterogeneity of variance. Once again, the limited number of effect sizes within these two variables prevents any search for moderators to resolve this heterogeneity.
Individual constructs displaying significant associations with PA were further analysed using Nagelkerke's R to determine the amount of variance accounted for in PA. The presence of heavy traffic explained the least amount of variance (R2 = 0.04), PA facilities and sidewalks explained 5% and 6% of physical activity variance respectively, whilst shops and services explained the greatest amount, accounting for 7%.
Discussion
The purpose of this review was to quantitatively summarize the associations between selected perceived environment variables and PA from individual studies. Our findings confirm previous suggestions that the perceived environment has a modest, yet significant association with PA [54,55]. This is evidenced by individual variables explaining relatively small amounts of variance (4–7%) in PA; however, the contribution of these potential changes to community behavior may be great. Since people living in a particular environment can be influenced by that setting [55], favorable alterations to communities may produce small changes in behaviors of entire populations [54]. Therefore, identifying and modifying environments to produce positive changes in PA are important. Individual level interventions promoting initial changes in PA can be complimented by interventions creating activity friendly environments, to assist in maintaining positive changes in PA. Therefore, multilevel interventions targeting individuals and communities are likely to be the most effective in changing PA.
Reporting the presence of proximal PA facilities, sidewalks, shops and services and that traffic was not a problem were all positively associated with PA. Although previous narrative reviews [11,19] were unable to provide consensus concerning how perceptions of traffic influence PA, the current review found positive associations across studies between the absence of traffic and PA using adjusted ORs. Since it has been noted that many road systems are designed without the needs of pedestrians in mind [56], it is time for planners to recognize the health relevance of their work. For instance, providing sidewalks separated from roads, pedestrian refuge islands, increased street lighting and roundabouts are all highly effective methods of reducing pedestrian crashes [56]. Engineering modifications can reduce traffic volumes and speeds if implemented effectively [57]. Although the potential for road modifications to reduce traffic volumes and increase individual PA may be small in magnitude, such modifications may contribute to sustainable positive changes in PA levels across the community.
Reviews from the transportation domain suggest the environment provides both cues and opportunities for people to engage in PA within their neighborhoods [5,6,58]. The current review supports such claims and provides evidence that these relationships exist across different populations. A plausible mechanism for observed associations is that neighborhood footpaths increase recreational and utilitarian walking/cycling by reducing the risk of falls (by providing people with opportunities to walk on a smooth, flat surfaces) and increasing perceptions of safety from traffic (due to barriers between roads and sidewalks) [59]. Sidewalk provision and town planning are under the control of various levels of government – local, state, federal. The various levels of government are capable of providing legislation that will promote or restrict community capacity to provide destinations (both utilitarian and recreational) and infrastructure (sidewalks and safe travel routes for non-motorized transportation) in neighborhood areas, influencing community behavior accordingly. The effects of such policy-level changes for changing PA in entire communities are potentially large and effective. Policy modifications are likely to influence PA by mandating the provision of safe environments close to the home (from both traffic and crime), useable green space and other recreational locations in which to engage in PA close to the home, active transport routes separate from vehicle traffic, and increases in school PE. Despite the difficulty of effecting change through policy or 'distal leverage points' [9], these and similar changes will likely contribute to longer lasting changes in behavior by making PA an easier "choice". Research examining how policy influences behavior [45,60,61] is promising, however research examining wider policy influences is needed to identify the most effective in creating environments producing sustainable increases in PA.
Because few studies have used objective measures of PA, the current review was limited to studies using measures of perceived environment. Although many of these studies used reliable measures of PA, such as the BRFSS [42,45], the lack of precise measurement of PA and environmental constructs may be obscuring true associations between PA and environmental characteristics [54]. Future studies are therefore encouraged to use both self-report and objective measures of PA [62], and when possible, combine self-report and objective measures of the environment to improve the predictive ability of studies. In addition, relatively few of included studies were conducted outside of the US, therefore research examining how environments influence activity in other countries is strongly encouraged.
Environmental changes may have differential effects on various sub-groups of the population (i.e. men and women may respond differently to similar aspects of the environment or environmental changes) [63]. The current review, limited to non-experimental studies, cannot answer these questions. As such, it is recommended that future research adopt a quasi-experimental approach to examine if similar environmental changes influence behaviors separately in regions where key socio-demographic measures are substantially different.
Some characteristics identified in this study, while likely difficult to alter, may be more readily manipulated than existing land densities and land-use mixes, which have received much attention within transportation literature [58]. Comparison studies demonstrate that rates of walking and cycling were higher in neighborhoods classified as transit orientated (higher connectivity) compared to neighborhoods classified as automobile orientated (lower connectivity) [64]. Additionally, modeling activities demonstrate that providing innovative block-cutting passages can increase connectivity of neighborhoods previously considered unfavorable to neighborhood walking [65]. Therefore, the provision of footpaths connecting previously unconnected neighborhoods and increasing connectivity may be a viable way to facilitate increased neighborhood PA. However, caution should be used in their planning and design to properly address safety and security concerns by providing good lighting and allowing residents and other street users' lines of sight into block-cutting passages. These modifications are effective strategies in reducing neighborhood crime [66,67].
The current research had several limitations that should be considered when examining the results. First, the sample was limited to studies that were published or accepted for publication, included adult populations, and used logistic regression analysis to determine associations between the perceived environment and activity. The limit to adult populations was imposed because very few studies examined features of the built or perceived environment in relation to the PA of children and youth. Thus, more research should be conducted with children and youth.
Second, two variables (PA facilities, presence of shops and services) displayed significant heterogeneity of variance suggesting that those groups of effect sizes do not represent a common population and that potential sources of heterogeneity should be sought. Unfortunately, the limited number of studies included in the analysis prohibited any search for moderators in the sample and thus the results for these two variables should be interpreted with caution. Likely sources of heterogeneity are the measurement instruments used to assess PA and the types of PA that were assessed in the original studies (e.g., walking, sufficient level of total PA). The possibility of PA measures and outcome measures as sources of heterogeneity should be examined in future analyses. Third, while several self-report measures with acceptable reliability [68,69] are available, studies used a variety of different measures to examine the perceived environment. Since alterations to measures may limit the possibility of making direct comparisons to previous studies, future research should use the most reliable measures in their entirety. Fourth, due to imprecise measurement of environmental constructs, variables were only included if a minimum of five ES were present. The number of ESs (n = 5) used in the analysis of outcome measures in previous meta-analyses [34,70] are similar to that used in the current study. However these studies used outcome measures (stroke, CVD) able to be measured with greater accuracy, and included larger within study sample sizes in pooled studies [34,70]. This minimum criterion may have reduced the number of constructs examined, however this delimitation was adopted to increase the confidence in ES estimation. Finally, all included studies were cross-sectional in design. Such designs may be subject to limitations including participants' self-selecting neighborhoods displaying design characteristics attractive to their own travel behaviors, attitudes [71], and PA preferences. For instance, using a pre-test/post-test study design, Krizek (2000) found that one-third of participants moved relatively short distances (<4 km) to neighborhoods possessing similar design characteristics, suggesting that neighborhood choice is based primarily on individual preference for a particular type of neighborhood [71]. Such limitations likely influence the results of included studies, and should be considered when interpreting the current results.
This quantitative review provides an objective summary of the association between a number of perceived environment variables and PA. The results should assist researchers in designing future research in the area. The small proportions of PA variance explained by the constructs reviewed suggest a need for the application of more comprehensive ecological models that include demographic, psychosocial, environmental, and biological variables.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MD collated the literature, conducted the synthesis and meta-analysis and drafted the overall manuscript. JS assisted in the meta-analytic procedures and editing of the manuscript. KM conceived the study, and participated in its design and coordination. All authors read and approved the final manuscript.
Table 2 Crude and adjusted odds ratios of selected perceived environmental variables associations with physical activity.
Variable Crude OR (95%CI) Q Adjusteda OR (95%CI) Q
PA Facilities in Neighborhood
No 1.00 1.00
Yes 0.99 (0.98–1.00) Q(6) = 64.41* 1.20 (1.06–1.34) Q(7) = 22.58*
Sidewalks/Footpaths Present
No 1.00 1.00
Yes 0.99 (0.97–1.01) Q(6) = 8.47 1.29 (1.17–1.41) Q(8) = 7.04
Shops & Services in Walking Distance
No 1.00 1.00
Yes 1.05 (0.80–1.29) Q(3) = 3.81 1.30 (1.14–1.46) Q(6) = 25.22*
Heavy Traffic is Present
Yes 1.00 1.00
No 1.00 (0.86–1.14) Q(7) = 5.34 1.22 (1.08–1.37) Q(6) = 6.32
High Crime is Present
Yes 1.00 1.00
No 0.90 (0.75–1.04) Q(7) = 9.27 0.96 (0.80–1.11) Q(4) = 5.89
Street Lighting in Present
No 1.00 1.00
Yes 0.94 (0.79–1.10) Q(6) = 9.08 1.02 (0.90–1.15) Q(7) = 5.39
Unattended Dogs are a Problem
Yes 1.00 1.00
No 0.90 (0.78–1.02) Q(7) = 8.31 0.88 (0.75–1.02) Q(4) = 2.30
aOR adjusted for those variables in original study
* Q is significant at the 0.05 level
Acknowledgements
The first author acknowledges the financial support of the Research Training Scheme (RTS) from Central Queensland University in the development of the present paper.
==== Refs
Stephenson J Bauman A Armstrong T Smith B Bellow B The cost of illness attributable to physical inactivity in Australia: A preliminary study 2000 Canberra: The Commonwealth Department of Health and Aged Care and the Australian Sports Commission
Katzmarzyk PT Gledhill N Shephard RJ The economic burden of physical inactivity in Canada Cmaj 2000 163 1435 1440 11192648
Colditz GA Economic costs of obesity and inactivity Med Sci Sports Exerc 1999 31 S663 667 10593542 10.1097/00005768-199911001-00026
Owen N Leslie E Salmon J Fotheringham MJ Environmental determinants of physical activity and sedentary behavior Exerc Sport Sci Rev 2000 28 153 158 11064848
Jackson RJ Kochtitzky C Creating a healthy environment: The impact of the built environment on public health Sprawl Watch Clearinghouse Monograph Series Retrieved July 7, 2003 2003
Frank LD Engelke PO Schmid TL Health and Community Design: The impact of the built environment on physical activity 2003 Washington: Island Press
Ewing R Schmid T Killingsworth R Zlot A Raudenbush S Relationship between urban sprawl and physical activity, obesity, and morbidity Am J Health Promot 2003 18 47 6 13677962
Sturm R Cohen DA Suburban sprawl and physical and mental health Public Health 2004 118 488 496 15351221 10.1016/j.puhe.2004.02.007
Booth SL Sallis JF Ritenbaugh C Hill JO Birch LL Frank LD Glanz K Himmelgreen DA Mudd M Popkin BM Environmental and societal factors affect food choice and physical activity: rationale, influences, and leverage points Nutr Rev 2001 59 S21 39 discussion S57–65. 11330630
Sallis JF Owen N Glanz K, Rimer BK, Lewis FM Ecological models of health behavior Health Behavior and Health Education: Theory, Research and Practice 2002 3 San Francisco: Jossey-Bass 462 484
Owen N Humpel N Leslie E Bauman A Sallis JF Understanding environmental influences on walking: Review and research agenda Am J Prev Med 2004 27 67 76 15212778 10.1016/j.amepre.2004.03.006
Eyler AA Brownson RC Bacak SJ Housemann RA The epidemiology of walking for physical activity in the United States Med Sci Sports Exerc 2003 35 1529 1536 12972873
Giles-Corti B Donovan RJ Socioeconomic status differences in recreational physical activity levels and real and perceived access to a supportive physical environment Prev Med 2002 35 601 611 12460528 10.1006/pmed.2002.1115
King AC Castro C Wilcox S Eyler AA Sallis JF Brownson RC Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of U.S. middle-aged and older-aged women Health Psychol 2000 19 354 364 10907654 10.1037//0278-6133.19.4.354
Duncan M Mummery K Psychosocial and environmental factors associated with physical activity among city dwellers in regional Queensland Prev Med 2005 40 363 372 15530589 10.1016/j.ypmed.2004.06.017
Centers for Disease Control Neighborhood safety and the prevalence of physical inactivity-selected states, (1996) JAMA 1999 281 1373 1374 10217042 10.1001/jama.281.15.1373
Troped PJ Saunders RP Pate RR Reininger B Addy CL Correlates of recreational and transportation physical activity among adults in a New England community Prev Med 2003 37 304 310 14507486 10.1016/S0091-7435(03)00137-3
McCormack G Giles-Corti B Lange A Smith T Martin K Pikora TJ An update of recent evidence of the relationship between objective and self-report measures of the physical environment and physical activity behaviours J Sci Med Sport 2004 7 81 92 15214606
Humpel N Owen N Leslie E Environmental factors associated with adults' participation in physical activity: a review Am J Prev Med 2002 22 188 199 11897464 10.1016/S0749-3797(01)00426-3
Hunter JE Schmidt FL Methods of meta-analysis: correcting error and bias in research findings 1990 London: Sage Publications
Hedges LV Olkin I Statistical methods for meta-analysis 1985 NY: Academic Press
Davey Smith G Egger M Phillips AN Meta-analysis. Beyond the grand mean? Bmj 1997 315 1610 1614 9437284
Gotzsche PC Reference bias in reports of drug trials Br Med J (Clin Res Ed) 1987 295 654 656 3117277
Teagarden JR Meta-analysis: whither narrative review? Pharmacotherapy 1989 9 274 281 discussion 281-274 2682549
Ravnskov U Cholesterol lowering trials in coronary heart disease: frequency of citation and outcome Bmj 1992 305 15 19 1638188
Strube MJ Hartmann DP A critical appraisal of meta-analysis Br J Clin Psychol 1982 21 129 139 7126951
Egger M Smith GD Phillips AN Meta-analysis: principles and procedures Bmj 1997 315 1533 1537 9432252
Wilcox S Castro C King AC Housemann R Brownson RC Determinants of leisure time physical activity in rural compared with urban older and ethnically diverse women in the United States J Epidemiol Community Health 2000 54 667 672 10942445 10.1136/jech.54.9.667
Eyler AA Matson-Koffman D Young DR Wilcox S Wilbur J Thompson JL Sanderson B Evenson KR Quantitative study of correlates of physical activity in women from diverse racial/ethnic groups: The Women's Cardiovascular Health Network Project – summary and conclusions Am J Prev Med 2003 25 93 103 14499815 10.1016/S0749-3797(03)00170-3
Humpel N Owen N Iverson D Leslie E Bauman A Perceived environment attributes, residential location, and walking for particular purposes Am J Prev Med 2004 26 119 125 14751322 10.1016/j.amepre.2003.10.005
Foster C Hillsdon M Thorogood M Environmental perceptions and walking in English adults J Epidemiol Community Health 2004 58 924 928 15483308 10.1136/jech.2003.014068
Suminski RR Poston WS Petosa RL Stevens E Katzenmoyer LM Features of the neighborhood environment and walking by U.S. adults Am J Prev Med 2005 28 149 155 15710269 10.1016/j.amepre.2004.09.009
Paganini-Hill A Perez Barreto M Stroke risk in older men and women: aspirin, estrogen, exercise, vitamins, and other factors J Gend Specific Med 2001 4 18 28
Oguma Y Shindo-Tagawa T Physical activity decreases cardiovascular disease risk in women: Review and meta-analysis Am J Prev Med 2004 26 407 418 15165657 10.1016/j.amepre.2004.02.007
Gillum RF Mussolino ME Ingram DD Physical activity and stroke incidence in women and men. The NHANES I Epidemiologic Follow-up Study Am J Epidemiol 1996 143 860 869 8610699
Haddock CK Rindskopf D Shadish WR Using Odds Ratios as Effect Sizes for Meta-Analysis of Dichotomous Data: A Primer on Methods and Issues Psychol Methods 1998 3 339 353 10.1037//1082-989X.3.3.339
Lipsey MW Wilson DB Practical meta-analysis 2001 49 London: Sage Publications
Petitti DB Meta-Analysis, Decision Analysis, and Cost-Effectiveness Analysis 1994 NY: Oxford University Press
Nagelkerke NJD A note on a general definition of the coefficient of determination Biometrica 1991 78 691 692
Huston SL Evenson KR Bors P Gizlice Z Neighborhood environment, access to places for activity, and leisure-time physical activity in a diverse North Carolina population Am J Health Promot 2003 18 58 69 13677963
Pate RR Pratt M Blair SN Haskell WL Macera CA Bouchard C Buchner D Ettinger W Heath GW King AC Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine Jama 1995 273 402 407 7823386 10.1001/jama.273.5.402
Eyler AA Personal, social, and environmental correlates of physical activity in rural Midwestern white women Am J Prev Med 2003 25 86 92 14499814 10.1016/S0749-3797(03)00169-7
Eyler AA Matson-Koffman D Rohm Young D Wilcox S Wilbur J Thompson JL Sanderson BK Evenson KR Quantitative study of correlates of physical activity in women from diverse racial/ethnic groups: Women's Cardiovascular Health Network Project – introduction and methodology Am J Prev Med 2003 25 5 14 14499804 10.1016/S0749-3797(03)00159-4
Ainsworth BE Wilcox S Thompson WW Richter DL Henderson KA Personal, social, and physical environmental correlates of physical activity in African-American women in South Carolina Am J Prev Med 2003 25 23 29 14499806 10.1016/S0749-3797(03)00161-2
Brownson RC Baker EA Housemann RA Brennan LK Bacak SJ Environmental and policy determinants of physical activity in the United States Am J Public Health 2001 91 1995 2003 11726382
Evenson KR Sarmiento OL Tawney KW Macon ML Ammerman AS Personal, social, and environmental correlates of physical activity in North Carolina Latina immigrants Am J Prev Med 2003 25 77 85 14499813 10.1016/S0749-3797(03)00168-5
Humpel N Owen N Leslie E Marshall AL Bauman AE Sallis JF Associations of location and perceived environmental attributes with walking in neighborhoods Am J Health Promot 2004 18 239 242 14748314
Rohm Young D Voorhees CC Personal, social, and environment correlates of physical activity in urban African-American women Am J Prev Med 2003 25 38 44 14499808 10.1016/S0749-3797(03)00163-6
Sanderson BK Foushee HR Bittner V Cornell CE Stalker V Shelton S Pulley L Personal, social, and physical environmental correlates of physical activity in rural African-American women in Alabama Am J Prev Med 2003 25 30 37 14499807 10.1016/S0749-3797(03)00162-4
Thompson JL Wolfe VK Wilson N Pardilla MN Perez G Personal, social, and environmental correlates of physical activity in Native American women Am J Prev Med 2003 25 53 60 14499810 10.1016/S0749-3797(03)00165-X
Wilbur J Chandler PJ Dancy B Lee H Correlates of physical activity in urban Midwestern Latinas Am J Prev Med 2003 25 69 76 14499812 10.1016/S0749-3797(03)00167-3
Voorhees CC Rohm Young D Personal, social, and physical environmental correlates of physical activity levels in urban Latinas Am J Prev Med 2003 25 61 68 14499811 10.1016/S0749-3797(03)00166-1
Wilbur J Chandler PJ Dancy B Lee H Correlates of physical activity in urban Midwestern African-American women Am J Prev Med 2003 25 45 52 14499809 10.1016/S0749-3797(03)00164-8
Owen N Humpel N Salmon J Oja P Oja P, Borms J Environmental Influences on Physical Activity Health Enhancing Physical Activity 2004 6 Berlin: International Society of Sport Science and Physical Education
Spence JC Lee RE Toward a comprehensive model of physical activity Psychology of Sport and Exercise 2003 4 7 24 10.1016/S1469-0292(02)00014-6
Retting RA Ferguson SA McCartt AT A review of evidence-based traffic engineering measures designed to reduce pedestrian-motor vehicle crashes Am J Public Health 2003 93 1456 1463 12948963
Ewing R Impacts of Traffic Calming Transportation Quarterly 2001 55 33 45
Saelens BE Sallis JF Frank LD Environmental correlates of walking and cycling: findings from the transportation, urban design, and planning literatures Ann Behav Med 2003 25 80 91 12704009 10.1207/S15324796ABM2502_03
Pikora T Physical environmental factors in the local environment that may influence walking and cycling for recreation and for transport among sedentary workers and home makers PhD 2003 Perth: The University of Western Australia
Sharpe PA Granner ML Hutto B Ainsworth BE Association of environmental factors to meeting physical activity recommendations in two South Carolina counties Am J Health Promot 2004 18 251 257 14748316
Prodaniuk TR Plotnikoff RC Spence JC Wilson PM The influence of self-efficacy and outcome expectations on the relationship between perceived environment and physical activity in the workplace Int J Behav Nutr Phys Act 2004 1 7 15169560 10.1186/1479-5868-1-7
Telford A Salmon J Jolley D Crawford D Reliability and validity of a self-report and proxy-report physical activity questionnaire for children: the Children's Leisure Activities Study Survey Ped Exerc Sci 2004 16 64 78
Humpel N Marshall AL Leslie E Bauman A Owen N Changes in neighborhood walking are related to changes in perceptions of environmental attributes Ann Behav Med 2004 27 60 67 14979864 10.1207/s15324796abm2701_8
Cervero R Gorham R Commuting in transit versus automobile neighborhoods Journal of the American Planning Association 1995 61 210 226
Aultman-Hall L Roorda M Baetz BW Using GIS for evaluation of neighbourhood pedestrian accessibility Journal of Urban Planning and Development 1997 March 10 17
Painter K The influence of street lighting improvement on crime, fear and pedestrian street use after dark Landscape and Urban Planning 1996 35 193 201 10.1016/0169-2046(96)00311-8
Farrington DP Welsh BC Effects of improved street lighting on crime: a systematic review 2002 National Crime Prevention Council
Evenson KR Eyler AA Wilcox S Thompson JL Burke JE Test-retest reliability of a questionnaire on physical activity and its correlates among women from diverse racial and ethnic groups Am J Prev Med 2003 25 15 22 14499805 10.1016/S0749-3797(03)00160-0
Brownson RC Chang JJ Eyler AA Ainsworth BE Kirtland KA Saelens BE Sallis JF Measuring the environment for friendliness toward physical activity: a comparison of the reliability of 3 questionnaires Am J Public Health 2004 94 473 483 14998817
Lee CD Folsom AR Blair SN Physical activity and stroke risk: a meta-analysis Stroke 2003 34 2475 2481 14500932 10.1161/01.STR.0000091843.02517.9D
Krizek KJ A pre-test/post-test strategy for researching neighbourhood-scale urban form and travel behaviour J Transport Res Board 1722 2000 48 55
|
16138933
|
PMC1236952
|
CC BY
|
2021-01-04 16:38:04
|
no
|
Int J Behav Nutr Phys Act. 2005 Sep 5; 2:11
|
utf-8
|
Int J Behav Nutr Phys Act
| 2,005 |
10.1186/1479-5868-2-11
|
oa_comm
|
==== Front
Immun AgeingImmunity & ageing : I & A1742-4933BioMed Central London 1742-4933-2-121611532510.1186/1742-4933-2-12ReviewLife and death of lymphocytes: a role in immunesenescence Gupta Sudhir [email protected] Houfen [email protected] Ruifen [email protected] Sudhanshu [email protected] Sastry [email protected] Laboratories of Cellular and Molecular Immunology and Molecular Biology, Division of Basic and Clinical Immunology, University of California, Irvine, California 92697, USA2005 23 8 2005 2 12 12 20 4 2005 23 8 2005 Copyright © 2005 Gupta et al; licensee BioMed Central Ltd.2005Gupta 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.
Human aging is associated with progressive decline in immune functions, increased frequency of infections. Among immune functions, a decline in T cell functions during aging predominates. In this review, we will discuss the molecular signaling in two major pathways of apoptosis, namely death receptor pathway and mitochondrial pathway, and their alterations in both T and B lymphocytes in human aging with a special emphasis on naïve and different memory subsets of CD8+ T cells. We will also discuss a possible role of lymphocyte apoptosis in immune senescence.
Tumor necrosis factorFasmitochondriaagingmemory T cellsActivation-induced cell death
==== Body
Introduction
Apoptosis is a physiological form of cell death, which plays an important role in embryogenesis, metamorphosis, cellular homeostasis, tissue atrophy and removal of tumor and mutated cells. In the immune system, apoptosis appears to play a crucial role in selection of T cell repertoire in the thymus, deletion of self-reactive T lymphocytes and B lymphocytes, regulation of immunological memory, deletion of effector T cells following an effective immune response, and in the cytotoxicity of target cells by CD8+ T cells and natural killer cells [1-3]. There are two major signaling pathways of apoptosis (Figure 1), the death receptor pathway (extrinsic pathway) and intrinsic pathway the mitochondrial pathway [4-11]. The apoptosis via both pathways is mediated by the activation of a series of cysteine proteases, the caspases. Caspases act as molecular chainsaw, which cleave a number of cytoplasmic and nuclear substrates to induce characteristic of apoptosis. Although both pathways of apoptosis involve activation of common effector or executioner caspases, they differ in the activation of apical or initiator caspases. Caspases are present in inactive form as prozymes. Apical caspases are autolytically activated by homodimerization without undergoing cleavage, whereas executioner caspases are activated via cleavage of their prodomain by apical caspases. Both pathways also recruit different adaptor molecules. In this article we will review differential sensitivity of various T lymphocyte subpopulations to apoptosis and their changes during aging and the role of subsets of T cells that are sensitive or resistant to apoptosis in immune senescence. A role of apoptosis in B lymphocytes in aging will also be briefly discussed.
Figure 1 Two distinct pathways of apoptosis. Death receptor pathway and mitochondrial pathway use distinct initiator caspases but common effector caspases. Death receptor and mitochondrial pathways are linked via Bcl-2 family protein Bid.
Death Receptor Pathway of Apoptosis
Death receptors belong to a large family of tumor necrosis factor receptors (TNFRs) and nerve growth factor receptors (NGFRs). Following interaction with death receptor ligand the cytoplasmic death domain (DD) of death receptor undergo trimerization, which leads to recruitment of a set of adaptor proteins and proximal caspase to form a death-inducing signaling complex (DISC). DISC serves as a platform for the activation of downstream caspases and apoptosis. In the DISC, initiator caspases undergo activation by homodimerization and without cleavage. Activated initiator caspases cleave effector caspases, which cleaves a number of cytoplasmic and nuclear substrates to induce apoptosis. We will discuss three distinct forms of death receptor-mediated apoptosis, which have been studied in human aging.
Activation-induced cell death
The activation-induced cell death (AICD), in which activation of T cells occurs through proper engagement of T cell receptors (TCRs) by specific antigen bound to MHC molecule and influenced by antigen concentration, and co-stimulatory signals. AICD plays an essential role in both central and peripheral deletion (clonal deletion) events involved in tolerance and homeostasis [12]. The AICD appears to be mediated primarily by an interaction between CD95 and CD95L [13-15]. In the AICD, cells are initially activated by anti-CD3 for 5 days and then re-stimulated with anti-CD3 to induce apoptosis, whereas in CD95-mediated apoptosis cells are first activated with anti-CD3 and cultured in IL-2 containing medium followed activation with anti-CD95 antibody or CD95L to induce apoptosis. AICD occurs only in the cells of the immune system, whereas CD95-mediated apoptosis may occur in any cell type. CD95-CD95L interaction is essential for AICD in mature T cells in vitro [16,17] and in vivo for peripheral T cell deletion [18,19].
CD95-mediated apoptosis
CD95 is a member of type I transmembrane receptors that is constitutively expressed on lymphocytes; however, CD95L, a type II transmembrane protein is lacking from resting lymphocytes and is transcrptionally regulated and induced upon activation of lymphocytes. The steps of CD95-mediated apoptosis signaling pathway are shown in Figure 2. Upon ligation with soluble CD95L or anti-CD95 monoclonal antibodies CD95 undergoes trimerization. Cytoplasmic DD of CD95 recruits an adapter protein, the fas-associated death domain (FADD), which contain a death effector domain (DED). FADD then recruits and through homologous and protein-protein interaction binds to procaspse-8 (Flice) to form a death-inducing signaling complex (DISC), which serves as a platform to initiate enzymatic activation of apoptotic pathway. Procaspase-8 is autolytically activated by homdimerization to generate active caspase-8, which is released from the DISC into the cytoplasm where it cleaves effector caspases (caspase-3, caspase-6, caspase-7) to generate active effector caspases. Active effector caspases in turn cleave a number of substrates to elicit characteristic morphological and biochemical features of apoptosis. This classical pathway occurs in so called type I cells [20]. In type II cells, procaspases-8 levels are very low and therefore caspase cascade is amplified via mitochondrial pathway. Caspase-8 cleaves the Bid, a Bcl-2 family member, to produce a truncated form of Bid (tBid), which then translocates from the cytoplasm to the mitochondria and exerts is proapoptotic effect by inhibiting Bcl-2/Bcl-xL resulting in the release of cytochrome c, activation of intiator caspase-9 and then of effector caspases resulting in apoptosis [21].
Figure 2 CD95-mediated Apoptosis. CD95 upon ligation with CD95 ligand (CD95L) undergo trimerization resulting in the recruitment of fas-associated death domain (FADD) and procaspase-8 to form death-inducing signaling complex (DISC). Procaspase-8 is autolytically activated by homodimerization and released from the DISC into the cytosol, where it cleaves and activate effector caspases to induce apoptosis.
TNFR-mediated Apoptosis
TNF-α is a plieotropic cytokine, which exerts its biological activity by binding to both type I and type II receptors (TNFR-I and TNFR-II) and activating several signaling pathways [2-7,22-25]. TNFRs belong to a family of TNFRs/NGFRs [26]. Both TNFRs receptors contain one to five cysteine-rich repeats in their extracellular domains; however differ in their cytoplasmic domain. TNFR-1 contains DD whereas TNFR-2 lacks DD. Therefore, TNFR-I signals both cell survival and cell death signals; whereas TNFR-II primarily mediates primarily a cell survival signals. However, recent data suggest that TNFR-II may also participate in apoptosis and may potentiate death signal mediated by TNFR-I. Both cell survival and cell death signals mediated by TNFRs require distinct sets of adapter and other downstream signaling molecules.
Steps of TNFR-mediated signaling are shown in Figure 3. TNFR-I undergo trimerization of its receptor death domains, which in turn recruit an adaptor protein, TNFR-associated death domain (TRADD). TRADD then may recruit another adapter molecule, the Fas-associated death domain (FADD). FADD then recruits procaspase-8, which is autolytically activated and then induces apoptosis via activation of effector caspases. TRADD may recruit distinct sets of adapter proteins, TRAF-2 (TNF-R-associated factor-2) and receptor interactive protein (RIP). TRAF-2 and RIP stimulate pathways leading to activation NFκB. Studies in mice and humans have shown that NF-κB is a repressor of apoptosis [27-31]. However, until recently it was unclear how NF-κB activation by TNF-α could inhibit initiator caspase activation through the same receptor (TNFR-I).
Figure 3 TNFR-mediated apoptosis. TNFR-I upon interaction with TNF-α undergo trimerization and recruitment of TNFR-associated death domain (TRADD), TNFR-associated factor 2 (TRAF-2) and receptor-interacting protein (RIP) to form complex I. This complex activates NF-κB via phosphorylation of IKK and IκB. NF-κB inhibits apoptosis by inducing a number of ant-apoptotic molecules (Bcl-xL, cIAPs, FLIP, Gadd45β, A20). After internalization of TNFR-I, TRADD, TRAF 2, and RIP are dissociated from the complex and FADD and caspase-8 are recruited (Complex II) to induce apoptosis. TRAF2 also activate JNK and sustanin activation of JNK induces apoptosis via selective release of Smac from the mitochondria.
Recently, Jurg Tschopp's group has proposed a two complex model based upon their experimental findings that TNFR-I signaling involve assembly of two distinct complexes that sequentially activate NF-κB and caspases [32]. In this model, the binding of TNF to TNFR-I results in the formation (within minutes) of signaling complex I. This complex contains TNFR-I, TRADD, RIP, and TRAF-2. Signaling complex I leads to activation of NF-κB via recruitment of (IκB kinase) IKK complex and phosphorylation of IκB. The secondary complex is form possibly following TNFR-I internalization (>2 hours following interaction between TNF and TNFR-I) in which TRADD, RIP, and TRAF-2 dissociate from the receptor and recruits FADD and caspase-8 (complex II). In conditions of complex I signaling, which leads to strong NF-κB activation, gene expression of anti-apoptotic proteins is induced and the activation of initiator caspases in complex II is inhibited. In contrast, when complex I signaling results in weak or deficient NF-κB activation, the products of anti-apoptotic gene are not made, and complex II can signal apoptosis via activation of caspases.
A family of TRAFs functions as adaptor molecules for TNFR superfamily members by associating with the intracellular domain of these proteins and subsequently mediating downstream signaling events such as activation of NF-κB. TRAF2 is recruited to TNFR-I signal complex via TRADD and plays a positive role in canonical pathway that activates NF-κB through IKKβ. TRAF2 homodimers as well as TRAF1:TRAF2 heterodimers can associate with TNFR-II that is required for signaling and NF-kB activation [33] TRAF2 also plays a role in TNF-induced activation of JNK via MEKK1 [34]. TRAF2 also ubiquitinates RIP at K63 (without proteasomal degradation) to activate NF-κB. Unlike TNFR-I, TNFR-II binds TRAF2 directly, hence activates IKK and JNK (TRAF-2 is also involved in TNFR-II-mediated activation of NF-κB). TRAF-2 also recruits ancilliary proteins (cIAP1, cIAP2, TRAF1, A20) that modulate signaling though each TNFRs and inhibit apoptosis. cIAP-TRAF2 complex inhibits caspases-8 activation by an unknown mechanism. Simultaneous engagement of both TNFR-I and TNFR-II amplifies TNF-induced apoptosis [35,36]. This correlates with increased TNFR-II-induced degradation of TRAF2. Since TRAF2 recruits cIAPs to TNFR-I, its degradation by TNFR-II may facilitate apoptosis by dissociation of cIAP from TRAF-2-cIAP complex and therefore allowing activation of caspase-8. In addition, TRAF2 degradation may also attenuate TNFR-I-mediated activation of NF-κB and promote apoptosis.
Receptor-interactive protein (RIP) is serine/threonine kinase, which is a component of TNFR-I signaling complex and is required for TNFR-I-mediated NF-κB activation [37-39]. RIP contains three domains, including an N-terminal kinase domain, an intermediate domain (which interact with the RING finger domain of TRAF-2) and an N-terminal DD. RIP interacts with TRADD through their respective DDs via protein-protein interaction. RIP family consist of five members, including RIP2, RIP3, RIP4, and recently described RIP5 [40-44]. All RIP kinases share significant similarities in their N-terminus kinase domain, but differ in their C-terminus domain. RIP, RIP2 and RIP4 are involved in the activation of NF-κB (42–44); RIP4 is also involved in JNK activation. Recently, it has been reported that RIP3 and RP5 are involved in TNFα-induced apoptosis [40,42]. RIP3 exerts its pro-apoptotic activity by activating caspases and/or by inhibiting RIP- and TNFR-1-induced NF-κB activation.
NF-κB mediates its repressor effect on apoptosis by inducing the expression of a number of anti-apoptotic genes including cIAPs, FLIP, TRAF-1, TRAF-2, Bcl-2, and Bcl-xL [30,31,45].
Inhibitor of apoptosis protein (IAP) family proteins, originally identified in the genome of baculovirus, has a key role in the negative regulation of apoptosis [46,47]. The cIAP-1 and cIAP2, two structurally homologous proteins, belong to a family of death inhibitors sharing a motif found in a Baculovirus inhibitor of death. cIAP1 and cIAP2 were initially isolated by their interaction with TRAF-1 and TRAF-2 in the TNF-RII complex. cIAP1 is also recruited to the DISK of TNF-RI by TRAF-2. In addition to cIAP1 and cIAP2, XIAP have a conserved COOH-terminal RING finger, zinc-binding domain [48]. Overexpression of these mammalian IAPs confers resistance to apoptosis. These proteins suppress apoptosis by preventing the activation of procaspases and inhibiting directly the enzyme activity of mature caspases. XIAP is a potent, active site-directed inhibitor of the effector caspases-3. In addition, TRAF-2-IAP complex inhibits caspases-8 activation by an unknown mechanism.
A20, a ring finger protein, was initially characterized as an inhibitor of TNF-α-induced apoptosis [49]. A20 is peculiar because it has dual activity in that it inhibits apoptosis as well as NF-κB activation [50]. These activities of A20 are cell type specific. A20 inhibits NF-κB activation by both deubiquitination (of K63 ubiquitination of RIP) and subsequent K48 ubiquitination for S26 proteasomal degradation of RIP. The fact that the expression of A20 is itself under control of NF-κB suggests that A20 is involved in the negative feed-back regulation of NF-κB activation. In contrast, A20 inhibits apoptosis, at least partially, by binding to TXBP151, which inhibits TNF-α-induced apoptosis. Furthermore, A20 and cIAP interact with a common region in TRAF2 [51]. Therefore, it is possible that A20 releases cIAP from the TRAF2-signaling complex, thereby allowing these proteins to exert their anti-apoptotic effects. Anti-apoptotic activity of A20 is restricted to certain cell type and is associated with decreased activation of caspases-3.
cFLIP is one of the apoptosis regulatory molecules that is induced by NF-κB [52]. FLIP comes in two spliced forms, the c-FLIPL and c-FLIPs. c-FLIPs contains two tandem repeat death effector domains (DED) and inhibits procaspase activation in the DISC. In contrast, c-FLIPL shares extensively homology with procaspase-8 yet it is enzymatically inactive [53]. In addition to its inhibitory effect on procaspase-8 activation, c-FLIP associates with Raf-1, which activates MEK1 to activate ERK, and with TRAF1 and TRAF2, which lead to NF-κB activation [54].
MAPK may inhibit [55] or promote apoptosis [56] via transient (inhibits apoptosis) or sustained (promotes apoptosis) activation of Janus-like kinase (JNK). Recently, a role of JNK in TNF-induced apoptosis has been explored [57]. JNK activation is required for TNF-induced apoptosis. Deng et al [58] demonstrated that TNF-α-induces apoptosis via sustained activation of JNK, which cleaves Bid, in a caspases-8-independent manner, to yield a unique 21kDa Bid cleaved product (jBid), which is different from caspases-8-dependent cleaved Bid (tBid) of 15kDa. jBid translocates to the mitochondria and preferentially releases Smac/Diablo from the mitochondria, which may disrupt TRAF-2-cIAP1 complex formation and its inhibition on caspases-8 activation. In addition Smac inhibits anti-apoptotic effects of cIAP and XIAP by binding it to them. De Smaele et al [59] identified GADD45β as an inhibitor of JNK activation and inhibitor of TNF-α-induced apoptosis. However, gadd45β is the only gene in this family that appears to be regulated by NF-κB and its ectopic expression completely suppresses TNF-α-induced apoptosis. This provides another mechanism via which NF-κB inhibits apoptosis.
Unlike TNF-RI, TNF-RII lack a cytoplasmic DD, instead interaction between TNF-α and TNF-RII results in binding of TRAF1 and TRAF2 to the cytoplasmic portion of TNF-RII. This then recruits the cellular inhibitor of apoptosis proteins cIAP-1 and cIAP-2 [46,51]. However, it has been reported that TNF-RII may also play an important role in the regulation of apoptosis through TNF-RI. Several investigators have reported that TNF-RII potentiates TNF-α-induced apoptosis [60-64] and proposed a number of mechanisms to explain this observation, including TNF-RII serving as high affinity trap of TNF-α that delivers TNF-α to TNF-RI [65], and direct induction or potentiation of apoptosis by the cytoplasmic domain of TNFR II [62,66].
Mitochondrial Pathway of Apoptosis
Several recent publications have reviewed the subject of mitochondrial pathway of apoptosis [7-11,67]. A number of stimuli, including chemotherapeutic agents, UV radiation, stress molecules (reactive oxygen and reactive nitrogen species) and growth factor withdrawal may mediate apoptosis via mitochondrial pathway In certain cell type mitochondrial pathway may provide an amplifying mechanism for death receptor-mediated apoptosis. Mitochondria contain two well-defined compartments: the matrix, surrounded by the inner membrane (IM), and the intermembrane space, which is surrounded by the outer membrane (OM). The IM contains various molecules, including ATP synthase, electron transport chain, and adenine nucleotide translocator (ANT). Under physiological conditions these molecules allow the respiratory chain to create an electrochemical gradient (membrane potential). The OM contains a voltage-dependent anion channel (VDAC). Bcl-2 is located on the IM and appears to play an important role in the maintenance of mitochondrial membrane potential (ΔΨm). The intermembrane space contains holocytochrome c, certain pro-caspases, adenylate kinase 2, Endo G, Daiblo/Smac, and apoptosis-inducing factor (AIF). The permeabilization of the OM, therefore, results in the release of these molecules into the cytoplasm. IM permeabilization leads to changes in ΔΨm. Once released from the mitochondria, cytochrome c binds to an adapter molecule Apaf-1 (Apoptotic protease-activating factor) in the presence of ATP/dATP and recruits pro-caspase 9 for form apoptosome (Fig. 4). Procaspase-9 is dimerized and activated without undergoing cleavage, and active caspases-9 activates executioner caspases to orchestrate apoptosis.
Figure 4 Mitochondrial pathway of apoptosis. See text for details.
A number of molecules present in the mitochondrial intermembrane space can promote apoptosis in caspases-independent manner. Htra2/Omi, in addition to its ability to block IAPs, appears to promote caspases-independent apoptosis via its serine protease activity [68,69]. Apoptosis inducing factor (AIF) is a caspases-independent death effector, which upon induction of apoptosis translocates from intermembrane space of the mitochondria to the nucleus where it AIF causes chromatin condensation and large scale DNA fragmentation [70,71]. Endo G, upon its release from mitochondrial intermembrane space, appears to directly mediate nuclear DNA fragmentation in a caspase-independent manner [72].
The mitochondrial membrane permeabilization (MMP) is controlled by a variety of members of the Bcl-2 family [7-11,73]. The Bcl-2 family members are divided into three groups: anti-apoptotic (Bcl-2, Bcl-xL, Mcl-1, Bcl-w, and A1), pro-apoptotic "BH3 only" (Bid, Bim, Bik, Bmf, Bad, Hrk, BNIP3) and pro-apoptotic "BH-123" (Bax, Bak, and Bok) proteins.
Several of the pro-apoptotic members of the Bcl-2 family, including Bax, Bak, Bad, Bid, and Bim, initiate MMP by forming what appears to be a channel. In order to influence their effects, the members of Bcl-2 pro-apoptotic family must dock onto the mitochondrial OM. During apoptosis Bax, which is present in the cytoplasm in a monomer form, is translocated to the mitochondrial membrane to form a dimer or high order oligomers. Bak can also loosely associate with OM. Bim, present in microtubules, also translocates to OM during apoptosis. Bim is a calcium-dependent proapoptotic molecule. Bcl-2 and Bcl-xL inhibit cytochrome C release. The phosphorylation of members of the Bcl-2 family rendered them inactive. In response to genotoxic agents, the stress-activated protein kinase (SAPK, also termed c-jun amino-terminal kinase or JNK) translocates to mitochondria and phosphorylates Bcl-xL, leading to Bcl-xL inactivation and induction of apoptosis.
Apoptosis in T Lymphocytes in Aging
Apoptosis in lymphocytes in aged humans has been studied primarily via death receptor signaling. Recently we (manuscript submitted) and others [74] have also studied apoptosis in human B lymphocytes.
Death receptor-induced apoptosis in CD4+ and CD8+ T cells in aging
During human aging (in contrast to mice) there is a progressive T cell lymphopenia, which is shared by both CD4+ and CD8+ T cells [75,76]. Although there has been controversy regarding lymphopenia in aging, our studies were performed in aged subjects from middle class social status, each of them own his/her house, living independently, and were asked to discontinue any anti-oxidants they might be taking for at least one week prior to the study (75). Therefore, our population of seniors does not have any nutritional or extrinsic factors and changes in lymphocyte counts and T cell subsets appear to reflect true changes of aging. Furthermore, many of our subjects were tested on two to three separate occasions. Although the precise mechanism of lymphopenia in aging is unclear, it is likely that decreased bone marrow precursors, decreased thymic output, reduced proliferative potentials and/or increased apoptosis, may contribute to T cell lymphopenia during human aging.
Activation-induced cell death (AICD) and CD95-mediated apoptosis in CD4+ and CD8+ T cells in aging
Apoptosis of T cells is increased during human aging [77-88]. Phelouzat et al [84,85] and Lechner et al [86] reported that T cells from aged individuals undergo increased AICD as compared to cells from young subjects and increased apoptosis was associated with increased expression of CD95. Potestio et al [87] reported increased spontaneous and AICD in T cells from aged humans and a correlation between increased spontaneous apoptosis and increased CD95 expression; however, we have observed better correlation between spontaneous apoptosis and CD95L expression rather than with CD95 expression [89].
In our study, using different methods to detect apoptosis including propidium iodide and TUNEL assay, Hoechst 33342 staining, and DNA fragmentation by gel electrophoresis, we observed that both CD4+ and CD8+ T cells from aged healthy subjects were more sensitive to anti-CD95-induced apoptosis as compared to young healthy control [77]. Increased apoptosis was associated with increased expression (at protein level) and increased and early activation of both caspase-8 and caspase-3 [90]. Furthermore, both CD4+ and CD8+ T lymphocytes from aged humans display increased expression of CD95 and CD95L. In addition, we observed higher apoptosis in CD4+ T cells as compared to CD8+ T cells. Zeng et al [91] have also observed preferential anti-CD95-induced death of CD4+ T cells.
TNFR-mediated apoptosis in CD4+ and CD8+ T cells
During aging, TNF-α production is increased [92-98]. We showed that both CD4 and CD8 cells from the elderly were more susceptible to TNF-α-induced apoptosis as compared to young subjects [2,6,7,76,78-82]. Furthermore, increased sensitivity of T cell subsets from aged humans to TNF-α-induced apoptosis was associated with increased and early activation of both caspase-8 and caspase-3. In contrast to our observations, Salvoni et al [99], using freshly isolated T cell subsets and using TNF-α and cyclohexamide to induce apoptosis, observed that aged CD4+ T cells were more resistant to TNF-α-induced-apoptosis as compared to young controls. However, these investigators demonstrated increased susceptibility of aged CD8+ T cells to apoptosis by Annexin V staining. In this study the expression of TNFRs or activation of caspases were not studied. These differences may be due to differential expression of TNFRs. The externalization of posphatidyl serine (which binds to Annexin V) is mediated by scramblase enzyme, which is sensitive to calcium. Therefore, significant changes in intracellular calcium may result in a cell to be positive for Annexin V without undergoing apoptosis; calcium signaling is different among CD4+ and CD8+ T cells and among young and aged T cells (unpublished data). In addition, no data of the effect of cyclohexamide alone or on Annexin V positivity was presented. In our study, we have used a model of in vivo activation and no cyclohexamide was used. The sensitivity of T cells to TNF-α-induced apoptosis appears to be age-dependent as cord blood lymphocytes are least sensitive [100] whereas aged T cells are most sensitive to TNF-α-induced apoptosis [78].
We also examined a role of downstream signaling molecules in increased apoptosis in aged T cells. We observed increased expression of TRADD and FADD in lymphocytes from aged subjects both at the level of mRNA and protein [77,78]. However, the expression of RIP both at the mRNA level and the protein level in aged lymphocytes was similar to lymphocytes from young subjects [78].
We have also reported that aged T cell subsets are sensitive to anti-CD95-induced apoptosis [76]. Since, FADD is common conduit for both CD95- and TNFR-mediated apoptosis we examine a role of increased FADD expression on increased apoptosis in aging. T cells from aged humans transfected with dominant negative FADD resulted in decreased TNF-α-induced apoptosis to a level comparable to young T cells, whereas wild type FADD resulted in increased apoptosis in both young and aged T cells albeit to a greater extent in young T cells to a level comparable to aged T cells, establishing a role of increased FADD in increased apoptosis in aged T cells [101].
Furthermore, we investigated whether downregulation of NF-κB activation (an anti-apoptotic signal) may also play a role in increased TNF-α-induced apoptosis. We have observed decreased TNF-α-induced DNA-binding activity of NF-κB in lymphocytes from aged humans as determined by EMSA and recently developed ELISA assay [102]. To further define the molecular mechanism of decreased NF-κB activity, we examined the expression of phosphorylated IKKβ and IκB. T cells from aged humans expressed low levels of phosphorylated IKKβ and Iκ-B. Furthermore, overexpression of IKKβ in aged T cells resulted in an increased phosphorylation of Iκ-B and decreased TNF-α-induced apoptosis in aged T cells to a level comparable to T cells from young subjects. NF-κB mediates its antiapoptotic effect via induction/upregulation of a number of anti-apoptotic genes, including Bcl-2, Bcl-xL, cIAPs, FLIP, and Gadd45β [30,31,45]. We have previously reported that in aging expression of Bcl-2 and cIAP1 is decreased [77,103]. We also showed that overexpressed IKKβ-induced inhibition of increased apoptosis in aged lymphocytes was associated with an upregulation of Bcl-2 and cIAP2 [102]. cIAP2 expression is regulated by NF-κB and therefore decreased cIAP2 in aging would be consistent with decreased NF-κB activity. Previously we have reported that Bcl-2 expression (another anti-apoptotic target of NF-κB) was decreased in aging [77]. These observations provide evidence for an important role and mechanisms by which decreased NF-κB sensitizes aging T cells to increased TNF-α-induced apoptosis. Our observations of decreased NF-κB activity in aged T cells is in agreement with those reported by Whisler et al [104] and Pahlvani and Harris [105]. Trebilcock and Ponnappan [106] demonstrated decreased induction of NF-κB in response to PMA and TNF-α. These authors further suggested that decreased induction of NF-κB could be due to decreased proteosome-mediated degradation of IκB [107]. In summary, it appears that decreased NF-κB activation contributes to the increased sensitivity of aged T cells to TNF-α-induced apoptosis.
Naïve, Central Memory and Effector Memory T Cells
Naïve T cells following exposure to a viral antigen undergo clonal expansion followed by clearance of virus. This phase is followed by a phase of contraction during which virus-specific T cells undergo apoptosis, and then number of virus-specific T cells stabilized and remained as memory T cells [108,109]. The memory T cells display differential expression of adhesion molecules (CD62L) and chemokine receptors (CCR-7), which allow them to home into lymph nodes and non-lymphoid tissue and mucosal sites, and to respond to microbes at peripheral tissue sites [110,111]. Therefore, CCR7+ and CD62high T cells are found in lymph nodes, whereas CCR7- and CD62Llow are found in extranodal sites such as liver and lung [112,113]. Based upon these adhesion molecules and chemokine receptors, memory CD8+ T cells have been divided into "central memory" T cells for those that are found in lymphoid organs and "effector memory" T cells that are found in peripheral non-lymphoid tissues and mucosal sites [114-116]. These subpopulations of naïve, central and effector memory T cells are identified by a number of cell surface proteins [109,114-117]. Recently, we have further characterized these subsets of CD8+ T cells [118]. Naïve CD8+ T cells in addition to expression of CD45RA and CCR7 also express CD27 and CD28, whereas central memory (TCM) CD8+ T cells retain these cell surface antigens except CD45RA. Effector memory CD8+ T cells are further subdivided into three subsets. One subset of effector memory (TEM-1) is CCR7-CD45RA-CD28+, the second set of effector memory CD8+ T cells(TEM-2) is CCR7-CD45RA-CD28-), and the third set of effector memory CD8+ T cells (TEM-3/TEMRA) is CDCR7-CD45RA+ CD28-). Fig. 5 shows phenotypic characteristics of naïve and various memory CD8+ T cells in humans. Although generally it is considered that TEM-3/TEMRA subset is lacking from CD4+ T cells, we have observed a very small subset of TEM-3/TEMRA CD4+ T cells (1%), which is increased in aging (unpublished data). In analyzing data of Salusco et al [108], we also noticed a small population of TEM-3/TEMRA CD4+ T cells, which authors did not discuss in their results. During subsequent discussion, we will be using terminology TEM and TEMRA for two effector memory T cell subsets.
Figure 5 Phenotypically distinct five distinct subsets of CD8+ T cells, including naïve, central memory (TCM) and three type of effector memory (TEM-1, TEM-2, and TEM-3) cells.
Figure 6 Expression of phosphorylated IKKβ, Bcl-2, and Bax in CD8+CD28+ and CD8+CD28- T cell lines. CD8+CD28- T cells, which are resistant to TNF-α-induced apoptosis, expressed increased levels of Bcl-2 and phospho IKKβ and decreased levels of Bax.
Apoptosis in Naïve, Central Memory and Effector Memory CD8+ And CD4+ T Cells
Death-receptor-induced apoptosis in naïve and memory CD4+ and CD8+ T cells
Recently, we have examined relative sensitivity of naïve and various memory CD8+ T cell subsets to TNF-α-induced apoptosis [83,119]. Mononuclear cells were activated with anti-CD3 monoclonal antibody for 2 days, cultured in an IL-2 containing medium for an additional three days and then activated with TNF-α. Our data show that naïve and TCM CD8+ T cells were sensitive whereas TEM and TEMRA CD8+ T cells were resistant to TNF-α-induced apoptosis. Apoptosis profile correlated with the activation of caspase-8 and caspase-3. However, no correlation was observed between relative sensitivity of four CD8+ T cell subsets to TNF-α-induced apoptosis and the expression of TNFR-I or TNFR-II. Therefore, we examined a role of downstream signaling events, including phosphorylation of IκB and NF-κB activity following activation with TNF-α and the expression of Bcl-2 and Bax in CD8+ T cell subsets. CD8+ CD28+ T cell line (containing naïve and TCM) and CD8+ CD28- T cell line (containing TEM and TEMRA were kindly provided by Dr. Abbe Vallejo, University of Pittsburg) were stimulated with TNF-α and IκB phosphorylation was measured by Western blotting, using IκB phospho antibodies and NF-κB activity was measured by ELISA-based assay. The expression of Bcl-2 and phosphorylated IκB and NF-κB activity were higher, whereas the expression of Bax was lower in TEM and TEMRA CD8+ T cells as compared to naïve and TCM CD8+ T cells (Figure 6). These data suggest that signaling molecules downstream of TNFRs may be responsible for differential sensitivity among subsets of CD8+ T cells to TNF-α-induced apoptosis. We have also observed that similar to CD8+ T cells, naïve and TCM CD4+ T cells (TCM> naïve) are sensitive to TNF-α-induced apoptosis, whereas TEM and TEMRA CD4+ T cells are resistant to TNF-α-induced apoptosis [120].
Naïve, Central Memory and Effector Memory CD4+ And CD8+ T Cells in Aging
In aging, there is a significant reduction in naïve CD8+ T cells [76] and CD8+ CD28+ T cells, which contain both naïve and central memory CD8+ T cells [121]. In addition, there is an accumulation of CD8+CD28- T cells, which are oligoclonal and show characteristics of cellular senescence (i.e. short telomere length indicative of long replicative history), and increased IFN-γ production [122-127]. These CD8+ T CD28- cells are comprised of two subpopulations of effector memory CD8+ T cells [107], namely TEM and TEMRA CD8+ T cells. Our study shows a marked decrease in naïve and TCM CD8+ T cells and an increase in TEM and TEMRA CD8+ T cells [83]. Fagnoni et al [76] also observed an increase in primed CD8+CD28-CD45RA+ (equivalent to TEMRA) in aged humans.
Apoptosis of Naïve, Central Memory and Effector Memory T Cell Subsets in Aging
Activation-induced cell death (AICD)
Herndon et al [128] reported an increased AICD of naïve T (CD45RO-) T cells in aged humans and suggested its role in age-associated T cell deficiency. However, this study did not investigate apoptosis in memory T cells. Brezinska et al [121] have reported that AICD (as measured by DNA content and caspasese-3 activation) in CD8+CD28+ (containing naïve and TCM) and CD8+CD28- (containing TEM and TEMRA) was comparable between young and aging. However, data was presented from a single middle aged individual.
CD95-mediated apoptosis
In our initial study, we observed that in aged humans, both CD45RA+ (naïve) and CD45RO+ (memory) CD4+ and CD8+ T cells were more sensitive to anti-CD95-induced apoptosis as compared to young subjects [77]. In addition, CD45RO+ displayed greater sensitivity to anti-CD95-induced apoptosis as compared to CD45RA+ CD4+ and CD8+ T cells in both young and aged subjects. Miyawaki et al (129) also reported that healthy adult memory T cells are more susceptible to anti-CD95-induced apoptosis as compared to naïve T cells. We reported decreased expression of Bcl-2 in both CD4+ and CD8+ T cells from aged humans as compared to young subjects; however, we did not examine Bcl-2 expression in naïve and memory subsets [77]. Shinohara et al [130] demonstrated decreased Bcl-2 expression in memory subsets of CD4+ and CD8+ T cells in healthy adults. This would be consistent with our observation of increased sensitivity of memory T cell subsets to death-receptor-mediated apoptosis as compared to naïve T cell subsets. Although a role of Bcl-2 family protein in death receptor pathway has been argued, several investigators have demonstrated that Bcl-2 blocks anti-CD95-induced apoptosis in mitogen-activated T cells [131,132]. Therefore, it is likely that decreased Bcl-2 expression in aging may play a role in increased sensitivity of T cell subsets in aged humans. Since CD45RA+ (contain naïve and TEMRA) and CD45RA-/CD45RO+ (contain TCM and TEM) are heterogenous and display differential sensitivity (naïve and TCM are sensitive and TEM and TEMRA are resistant) to other death stimuli, further studies are warranted with CD95-mediated signal in naïve and different memory subsets of CD8+ T cells.
TNF-α-induced apoptosis
In our previous study we reported that both CD45RA+ naïve and CD45RA- memory CD4+ and CD8+ T cells from aged individuals were more sensitive to TNF-α-induced apoptosis [78]. Since CD45RA+ and CD45RA- T cells are heterogeneous we examined the relative sensitivity of naïve, TCM, TEM and TEMRA CD8+ and CD4+ T cell subsets to TNF-α-induced apoptosis in young and aged subjects. In aged humans, we observed that naïve and central memory CD8+ T cells displayed increased TNF-α-induced apoptosis as compared to young subjects, which is associated with increased caspase-8 and caspase-3 activation. Therefore, it appears that during aging decrease in naïve CD8+ T cells may be due to both decreased thymic output as well as increased apoptosis. We have also observed greater increased in apoptosis in TCM CD8+ T cells as compared to naïve CD8+ T cells in aged humans. In contrast, no significant difference was observed in the apoptosis of TEM and TEMRA CD8+ T cells between aged and young humans; both were comparably resistant to apoptosis [120]. This would suggest that the accumulation of TEM and TEMRA CD8+ T cells in aged humans is not due to changes in apoptosis and may be due to increased growth. We have observed that both TEM and TEMRA CD8+ T cells from young and aged subjects proliferate well in the presence of exogenous IL-2 and IL-15 even more than TCM CD8+ T cells (unpublished observation). We have also observed increased expression of IL-15 gene in CD8+ T cells from aged humans (by gene array) as compared to young subjects. These observations suggest that CD8+CD28- T cells generated by repeated activation in vitro are not a true model for CD8+CD28- T cells in aged humans since the latter cells do not proliferate (replicative senescence).
Since the expression of TNFR-I or TNFR-II is similar in young and aged humans, we have examined role of downstream signaling events in increased sensitivity of naïve and TCM CD8+ T cells in aged humans to TNF-α-induced apoptosis (manuscript in preparation). We have observed that CD28-CD8+ (containing naïve and TCM) from aged subjects display decreased phosphorylation of IKKα/β and IκB and decreased activation of NF-κB. Since NF-κB mediates its anti-apoptotic effect via induction of a number of anti-apoptotic molecules (IAP, FLIP, A20, Bcl-xL), we examined expression of these molecules by Western blotting. cIAP1, FLIPL, FLIPS, A20, and BCL-xL expression were decreased in aging CD28-CD8+ T cells. These data would suggest that decreased NF-κB activity may be central to increased sensitivity of naïve and TCM CD8+ T cells and perhaps of CD4+ T cells (since they also show similar profile of apoptosis in aging) to TNF-α-induced apoptosis.
B Cells Subsets in Human Aging
B-lineage cells following immunoglobulin (Ig) gene rearrangement to generate functional antigen receptor are released into the peripheral blood B cell pool as naïve B cells. After exposure to a T-dependent antigen, Naïve be cells differentiate via one of two different pathways. They can either differentiate into short-lived Ig secreting cells or they migrate to germinal center, where high-affinity antigen-specific B cells are selected and undergo proliferation, somatic hypermutation of Ig V-region genes, isotype switching and develop into long-lived memory B cells [133-135]. Although a number of cell surface markers have been used to identify memory B cells including lack of surface IgD expression and expression of membrane IgG and IgA [135], or as IgD-CD38- B cells [136], these markers identify only certain populations of memory B cells. Recently, CD27 has been identified as a key marker of memory B cells and CD27 signaling promotes the differentiation of memory B cells to Immunoglobulin-secreting plasma cells [137].
Aging is associated with both quantitative and qualitative changes in humoral immunity. These include decreased levels of IgM and increased levels of IgG and IgA, decreased B cell repertoire, decreased primary and secondary specific antibody response to vaccine antigens and changes in antibody affinity [138]. It has been demonstrated that CD27 expression increases with age; lacking in cord blood B cells and approximately 40% of adult B cells express CD27 antigen [137]. We have examined the proportions and numbers of naïve and memory B cells in thirty young and fifty aged subjects. Our data show that the proportion of CD27+CD19+ memory B cells is significantly increased whereas the proportion of CD27-CD19+ naïve B cells is significantly decreased. This may explain reduced specific antibody response to novel antigens and increased accumulation of somatic mutation of Ig variable region genes in aged humans [139]. When B cells were analyzed for the expression of CD38 to define activated and switchable B cells, no significant difference was observed between young and aged subjects. Our observations are in complete contrast to recent report by Chong et al [74], who observed decreased memory and increased naïve B cells in aged subjects. The reason for this discrepancy is unclear. Our aged subjects were of middle socio-economical class, in good health and living independently. Since majority of seniors are on a number of supplements, including anti-oxidants and vitamin A and E, which can modify immune functions and apoptosis, our subjects were asked to discontinue all supplements at least one week prior to blood draw. Therefore, our population did not have any nutritional or chemical compounding factors. Chong et al [74] also demonstrated that naïve B cells were more resistant to spontaneous apoptosis as compared to memory B cells.
One small subpopulation of B cells express CD5 antigen, a 67 kDa monomer, which was originally identified as a subset of T cells. CD5+ B cells express a limited repertoire of V genes, secrete IgM antibodies that often react with self antigens (autoantibodies), and appear to be self-renewing population. These cells are expanded in autoimmune diseases. Since aging is associated with autoimmunity we have analyzed CD5+ B cells in aged subjects. We observed no difference in the proportions and numbers of CD5+ B cells between aged and young subjects. Furthermore, we examined the expression of CD95 and apoptosis in these subsets. We have observed increased proportions of CD95+CD5+ cells in aging as compared to young controls; however, the expression of CD95 did not correlate with apoptosis, which was comparable in young and aged subjects (manuscript submitted).
In summary, increased apoptosis in naive and TCM CD8+ T cells in aging appears to play an important role in lymphopenia of naïve and TCM CD8+ T cells (83), which might be responsible for decline in T cell functions and increased susceptibility to viral infection and increased frequency of cancer in aging. Data of B cells in aging is conflicting and more in-depth analysis is needed.
Acknowledgements
The Work cited is in part supported by a grant from UPHS AG-18313
==== Refs
Krammer PH CD95's deadly mission in the immune system Nature 2000 407 789 795 11048730 10.1038/35037728
Gupta S Suicidal journey in the Fas (t) track Recent Res Dev Immunol 2000 2 11 19
Gupta S Molecular steps of cell suicide: An insight into immune senescence J Clin Immunol 2000 20 229 239 10939710 10.1023/A:1006653917314
Ashkanazi A Dixit VM Death receptors: signaling and modulation Science 1998 281 1305 1308 9721089 10.1126/science.281.5381.1305
Gupta S Molecular steps of death receptor and mitocondrial pathways of apoptosis Life Sci 2000 69 2957 2964 11758823 10.1016/S0024-3205(01)01404-7
Gupta S Molecular steps of TNF receptor-mediated apoptosis Curr Mol Med 2001 1 299 306 10.2174/1566524013363780
Gupta S Decision between life and death during TNF-induced signaling J Clin Immunol 2002 22 270 278 12405160 10.1023/A:1016089607548
Green DR Evan GI A matter of Life and Death Cancer Cell 2002 1 19 30 12086884 10.1016/S1535-6108(02)00024-7
Kroemer G Reed JC Mitochondrial control of cell death Nature Med 2000 6 513 519 10802706 10.1038/74994
Martinou J-C Green DR Breaking the mitochondrial barrier Nature Rev Mol Cell Biol 2001 2 63 67 11413467 10.1038/35048069
Zamzami N Kroemer G The mitochondrion in apoptosis: how pandora's box opens Nature Rev Mol Cell Biol 2001 2 67 71 11413468 10.1038/35048073
Green DR Droin N Pinkoski M Activation-induced cell death in T cells Immunol Rev 2003 193 70 81 12752672 10.1034/j.1600-065X.2003.00051.x
Dhein J Walczak H Baumler C Debatin KM Kramer PH Autocrine T cell suicide mediated by APO-1/(fas-CD95) Nature 1995 373 438 441 7530335 10.1038/373438a0
Brunner T Mogil RJ LaFace D Yoo NJ Mahboubi A Echeverri F Martin SJ Force WR Lynch DH Ware CF Green DR Cell autonomous Fas (CD95)/Fas ligand interaction mediates activation-induced apoptosis in T-cell hybridomas Nature 1995 373 441 444 7530336 10.1038/373441a0
Ju ST Panka DJ Cui H Ettinger R el-Khatib M Sherr DH Stanger BZ Marshak-Rothstein A Fas (CD95/FasL interactions are required for programmed cell death after T cell activation Nature 1995 373 444 448 7530337 10.1038/373444a0
Anderson MR Tough TW Davis-Smith T Braddy S Falk B Schooley KA Goodwin RG Smith CA Ramsdell F Lynch DH Fas ligand mediates activation-induced cell death in human T lymphocytes J Exp Med 1995 181 71 77 7528780 10.1084/jem.181.1.71
Mixter PF Russell JQ Budd RC Delayed kinetics of T lymphocyte anergy and deletion in lpr mice J Autoimmunity 1994 7 697 710 7534079 10.1006/jaut.1994.1055
Mogil RJ Radvanyi L Gonzalez-Quintial R Miller R Mills G Theofilopoulos AN Green DR Fas (CD95) participates in peripheral T cell deletion and associated apoptosis in vivo Int Immunol 1995 7 1451 1458 7495753
Renno T Hahne M Tschopp J MacDonald HR Peripheral T cells undergoing superantigen-induced apoptosis in vivo express B220 and upregulate Fas and Fas ligand J Exp Med 1996 183 431 437 8627156 10.1084/jem.183.2.431
Scaffidi C Fulda S Srinivasan A Friesen C Li F Tomaselli KJ Debatin KM Krammer PH Peter ME Two CD95 (Apo-1/Fas) signaling pathways EMBO J 1998 17 1675 1687 9501089 10.1093/emboj/17.6.1675
Li H Zhu H Xu C Yuan J Cleavage of Bid by caspase-8 mediates mitochondrial damage in the Fas pathway of apoptosis Cell 1998 94 491 501 9727492 10.1016/S0092-8674(00)81590-1
Screaton G Xu X-N T cell life and death signaling via TNF-receptor family members Curr Opin Immunol 2000 12 316 3222 10781406 10.1016/S0952-7915(00)00093-5
Thomas B Grell M Pfizenmaier K Scheurich P Identification of a 60-kDa tumor necrosis factor (TNF) receptor as the major signal transducing component in TNF responses J Exp Med 1990 172 1019 1023 2170559 10.1084/jem.172.4.1019
Darnay BG Aggarwal BB Early events in TNF signaling: a story of associations and dissociations J Leukocyte Biol 1997 61 559 66 9129204
Wallach D Boldin M Varfolomeev E Beyaert R Vandenabeele P Fiers W Cell death induction by receptors of the TNF family: towards a molecular understanding FEBS Lett 1997 410 96 106 9247131 10.1016/S0014-5793(97)00553-X
Locksley RM Kileen N Lenardo MJ The TNF and TNF receptor superfamilies: interacting mammalian biology Cell 2001 104 487 501 11239407 10.1016/S0092-8674(01)00237-9
Beg AA Baltimore D An essential role for NF-κB in preventing TNF-α-induced cell death Science 1996 274 782 784 8864118 10.1126/science.274.5288.782
Ghosh S May MJ Kopp EB NF-κB and rel proteins: evolutionarily conserved mediators of immune responses Annu Rev Immunol 1998 16 225 260 9597130 10.1146/annurev.immunol.16.1.225
Baldwin AS The NF-κB and IκB proteins: new discoveries and insights Annu Rev Immunol 1996 14 649 681 8717528 10.1146/annurev.immunol.14.1.649
Karin M Lin A NF-κB at the crossroads of life and death Nature Immunol 2002 3 221 227 11875461 10.1038/ni0302-221
Ghosh S Karin M Missing pieces in the NF-kB puzzle Cell 2002 109 S81 S96 11983155 10.1016/S0092-8674(02)00703-1
Micheau O Tschopp J Induction of TNF receptor I-mediated apoptosis via two sequential signaling complexes Cell 2003 114 181 190 12887920 10.1016/S0092-8674(03)00521-X
Rothe M Wong SC Henzel WJ Goeddel DV A novel family of putative signal transducers associated with the cytoplasmic domain of the 75 kDa tumor necrosis factor receptor Cell 1994 78 681 692 8069916 10.1016/0092-8674(94)90532-0
Rothe M Sarma V Dixit VM Goeddel DD TRAF2-mediated activation of NF-κB by TNF receptor 2 and CD40 Science 1995 269 1424 1427 7544915
Wajant H Pfizenmaieer K Scheurich P Tumor necrosis factor signaling Cell Death Differ 2003 10 45 65 12655295 10.1038/sj.cdd.4401189
Maeda S Chang L Li ZW Luo JL Leffert H Karin M IKKβ is required for prevention of apoptosis mediated by cell bound but not by circulating TNF-alpha Immunity 2003 19 725 737 14614859 10.1016/S1074-7613(03)00301-7
Hsu H Huang J Shu HB Baichwal V Goeddel DV TNF-dependent recruitment of the protein kinase RIP to the TNF receptor-1 signaling complex Immunity 1996 4 387 396 8612133 10.1016/S1074-7613(00)80252-6
Kelliher MA Grimm S Ishida Y Kuo F Stanger BZ Leader P The death domain kinase RIP mediates the TNF-induced NF-κB signal Immunity 1998 8 297 303 9529147 10.1016/S1074-7613(00)80535-X
Ting AT Pimentel-Muinos FX Seed B RIP mediates tumor necrosis factor receptor 1 activation of NF-κB but not Fas/APO-1-initiated apoptosis EMBO J 1996 15 6189 6196 8947041
Zha J Zhou Q Xu L-G Chen D Li L Zhai Z Shu H-B RIP5 is a RIP-homologous inducer of cell death Biochem Biophys Res Comm 2004 319 298 303 15178406 10.1016/j.bbrc.2004.04.194
McCarthy JV Ni J Dixit VM RIP-2 is a novel NF-κB activating and cell death-inducing kinase J Biol Chem 1998 273 16969 16975
Sun X Lee J Navas T Baldwin DT Stewart TA Dixit VM RIP3, a novel apoptosis-inducing kinase J Biol Chem 1999 274 16871 16875 10358032 10.1074/jbc.274.24.16871
Chen L Haider K Mondo M Cariappa A Rowitch D Pillai S Protein kinase C-associated kinase (PKK), a novel membrane associated ankyrin repeat-containing protein kinase J Biol Chem 2001 276 21737 31744 11278382
Meylan E Martinon F Thome M Gscwendt M Tschopp J RIP4 (DICK/PKK), a novel member of RIP kinase family, activates NF-κB, and is processed during apoptosis EMBO Rep 2002 3 1201 1208 12446564 10.1093/embo-reports/kvf236
Pahl HL Activators and target genes of Rel/NF-kB transcription factors Oncogene 1999 18 6855 6866 10.1038/sj.onc.1203239
Deveraux QL Reed JC IAP family ptoteins: suppressor of apoptosis Genes Dev 1999 13 239 252 9990849
Salvesen GS Duckett CS IAP proteins: blocking the road to death's door Nature Rev Mol Cell Biol 2004 3 401 410 12042762 10.1038/nrm830
Devereaux QL Takahashi R Salvesen GS Reed JC X-linked IAP is a direct inhibitor of cell-death proteases Nature 1997 388 300 304 9230442 10.1038/40901
Opipari AW JrHu HM Yabkowitz R Dixit VM The A20 zinc finger protein protects cells from tumor necrosis factor cytotoxicity J Biol Chem 1992 267 12424 12427 1618749
Heyninck K Beyaert R A20 inhibits NF-κB activation by dual ubiquitin-editing functions Trends Biochem Sci 2005 30 1 4 15653317 10.1016/j.tibs.2004.11.001
Rothe M Pan MG Henzel WJ Ayres TM Goeddel DV The TNFR2-TRAF signaling complex contains two novel proteins related to baculoviral inhibitor of apoptosis proteins Cell 1995 83 1243 1252 8548810 10.1016/0092-8674(95)90149-3
Micheau O Lens S Gaide O Alevizopolous K Tshopp J NF-κB signals induce the expression of c-FLIP Mol Cell Biol 2001 21 5299 5305 11463813 10.1128/MCB.21.16.5299-5305.2001
Irmler M Thome M Hahne M Schnieder P Hoffmann K Steiner V Bodmer J-L Schroter M Burns K Mattmann C Rimoldi D French LE Tschopp J Inhibition of death receptor signals by cellular FLIP Nature 1997 388 190 195 9217161 10.1038/40657
van Drogen F O'Rourke SM Stucke VM Jaquenoud M Neiman AM Peter M The caspases-8 inhibitor FLIP promotes activation of NF-kappaB and ERK signaling pathways Curr Biol 1997 10 640 648
Natoli G Costanzo A Ianni A Templeton DJ Woodgett JR Balsano C Levero M Activation of SAPK/JNK by TNF receptor 1 through a noncytotoxic TRAF-2-dependent pathway Science 1997 275 200 203 8985011 10.1126/science.275.5297.200
Ichijo N Nishida E Irie K Ten Dijke P Saitoh M Moriguchi T Tagaki M Matsumoto K Miyazono K Gotoh Y Induction of apoptosis by ASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways Science 1997 275 90 94 8974401 10.1126/science.275.5296.90
Vartfolmeev EE Askenazi A Tumor necrosis factor: an apoptosis JuNKie? Cell 2004 116 491 497 14980217 10.1016/S0092-8674(04)00166-7
Deng Y Ren X Yang L Lin Y Wu X A JNK-dependent pathway is required for TNF-α-induced apoptosis Cell 2003 115 61 70 14532003 10.1016/S0092-8674(03)00757-8
De Smaele E Zazzeroni F Papa S Nguyen DU Jin R Cong R Franzoso G Induction of gadd45β by NF-κB downregulates proapoptotic JNK signaling Nature 2001 414 308 313 11713530 10.1038/35104560
Declercz W Denecker G Fiers W Vandenabeele P Cooperation of both TNF receptors in inducing apoptosis: involvement of the TNF receptor-associated factor binding domain of the TNF receptor 75 J Immunol 1998 161 390 399 9647248
Haridas V Darnay BG Natrajan K Helle R Aggarwal BB Overexpression of the p80 TNFR leads to TNF-dependent apoptosis, nuclear factor-kappa B activation J Immunol 1998 160 3152 3162 9531270
Weiss T Grell M Siekienski K Muhlenbeck F Durkop H Pfizenmaier K Scheurich P Wajant H TNFR80-dependent enhancement of TNFR60-induced cell death is mediated by TNFR-associated factor 2 and is specific for TNFR60 J Immunol 1998 161 3136 3142 9743381
Chen FK Lenardo MJ A crucial role for p80 TNF-R2 in amplifying p60 TNF-R1 apoptosis signals in T lymphocytes Eur J Immunol 2002 30 652 660
Vandenabeele P Declercq W Vanhaesebroeck B Grooten J Fiers W Both TNF receptors are required for TNF-mediated induction of apoptosis in PC60 cells J Immunol 1995 154 2904 2913 7876558
Tartaglia L Pennica D Goddel DV Ligand passing: the 75-kDa tumor necrosis factor (TNF) receptor recruits TNF for signaling by the p55-kDa TNF receptor J Biol Chem 1993 268 18542 18548 8395508
Grell M Zimmermann G Gottfried E Chen CM Grunwald U Huang DC Wu Lee Y Durkop H Englemann H Scheurich P Induction of cell death by tumor necrosis factor (TNF) receptor 2, CD40, and CD30: a role of TNFR1 activation by endogenous membrane-anchored TNF EMBO J 1999 18 3034 3043 10357816 10.1093/emboj/18.11.3034
Gupta S Molecular signaling in death receptor and mitochondrial pathways of apoptosis Internat J Oncol 2003 22 15 20
Hegde R Srinivasula SM Zhang Z Wassell R Mukattash R Cilenti L DuBois G Lezebnik Y Zervos AS Fernandes-Alnemri T Alnemri ES Identification of Omi/HtrA2 as a mitochondrial apoptotic serine protease that disrupts inhibitor of apoptosis protein-caspase interaction J Biol Chem 2002 277 432 438 11606597 10.1074/jbc.M109721200
Suzuki Y Imai Y Nakayama H Takahashi K Takio T Takahashi R A serine protease, HtrA2, is released from the mitochondria and interacts with XIAP, inducing cell death Mol Cell 2001 8 613 621 11583623 10.1016/S1097-2765(01)00341-0
Lorenzo HK Susin SA Penninger J Kroemer G Apoptosis inducing factor (AIF): a physiologically old, caspases-independent effector of cell death Cell Death Diff 1999 6 516 524 10.1038/sj.cdd.4400527
Loeffler M Daugas E Susin SA Zamzani N Metivier D Nieminen A-L Brothers G Penninger JM Kroemer G Dominant cell death induced by extra-mitochondrially targeted apoptosis-inducing factor FASEB J 2001 15 758 767 11259394 10.1096/fj.00-0388com
Li LY Luo X Wang X Endonuclease G is an apoptotic DNAase when released from mitochondria Nature 2001 412 95 99 11452314 10.1038/35083620
Reed JC Double identity for protein of Bcl-2 family Nature 1997 387 773 778 9194558 10.1038/42867
Chong Y Ikematsu H Yamaji K Nishimura M Nebishima S Kashiwagi S Hayashi J CD27+ (memory) B cell decrease and apoptosis- resistant CD27- (naïve) B cell increase in aged humans: Implications for age-related peripheral B cell developmental disturbance Internat Immunol 2005 17 383 390 10.1093/intimm/dxh218
Gupta S Tumor necrosis factor-α-induced apoptosis in T cells from aged humans: a role of TNFR-I and downstream signaling molecules Exp Gerontol 2002 37 293 299 11772515 10.1016/S0531-5565(01)00195-4
Fagnoni FF Vescovini R Paserri G Bologna G Pedrazzoni M Lavagetto G Casti A Franceschi C Passeri M Sansoni Shortage of circulating naïve CD8+ T cells provides new insights on immunodeficiency in aging Blood 2002 95 2860 2868 10779432
Aggarwal S Gupta S Increased apoptosis of T cell subsets in aging humans: Altered expression of Fas (CD95), Fas ligand, Bcl-2, and Bax J Immunol 1998 160 1627 1637 9469419
Aggarwal S Gollapudi S Gupta S Increased TNF-α-induced apoptosis in lymphocytes from aged humans: changes in TNF-α receptor expression and activation of caspases J Immunol 1999 162 2154 2161 9973490
Gupta S Tumor necrosis factor-α-induced apoptosis in T cell subsets from aged humans. Receptor expression and downstream signaling events Exp Gerontology 2002 37 293 299 10.1016/S0531-5565(01)00195-4
Gupta S Chiplunkar S Kim C Yel L Gollapudi S Effect of age on molecular signaling of TNF-α-induced apoptosis in human lymphocytes Mech Ageing Dev 2003 124 503 509 12714259 10.1016/S0047-6374(03)00028-9
Gupta S Molecular and biochemical pathways of apoptosis in lymphocytes from aged humans Vaccine 2000 18 1596 1601 10689134 10.1016/S0264-410X(99)00492-2
Gupta S A road to ruins: An insight into immunosenescence Adv Cell Aging Gerontol 2003 13 169 185
Gupta S Molecular mechanisms of apoptosis in the cells of the immune system in human aging Immunol Rev 2005 15882349
Phelouzat MA Arbogast A Laforge T Quadri RA Proust JJ Excessive apoptosis of mature T lymphocytes is a characteristic feature of human immune senescence Mech Ageing Dev 1996 88 25 38 8804091 10.1016/0047-6374(96)01714-9
Phelouzat MA Laforge T Abrogast A Quadri RA Boutet S Proust JJ Susceptibility to apoptosis of T lymphocytes from elderly humans is associated with increased in vivo expression of functional fas receptors Mech Ageing Dev 1997 96 35 46 9223109 10.1016/S0047-6374(97)01883-6
Lechner H Amort M Steger MM Maczek C Grubeck-Lobenstein B Regulation of CD95 (Apo-1) expression and the induction of apoptosis of human T cells: changes in old age Int Arch Allergy Immunol 1996 110 238 243 8688670
Potestio M Caruso C Gervasi F Scialabba G D'Anna C Di Lorenzo G Balistreri CR Candore G Romano GC Apoptosis and aging Mech Ageing Dev 1998 102 221 237 9720654 10.1016/S0047-6374(98)00020-7
Dennett NS Barcia RN McLeod JD Age-associated decline in CD25 and CD28 expression correlate with an increased susceptibility to CD95 mediated apoptosis in T cells Exp Gerontol 2000 37 271 283 11772513 10.1016/S0531-5565(01)00193-0
Gupta S Aggarwal S Nguyen T Increased spontaneous apoptosis in T lymphocytes in DiGeorge anomaly Clin Exp Immunol 1998 113 65 71 9697985 10.1046/j.1365-2249.1998.00629.x
Aggarwal S Gupta S Increased activity of caspase-3 and caspase-8 during Fas-mediated apoptosis in lymphocytes from aging humans Clin Exp Immunol 1999 117 285 290 10444259 10.1046/j.1365-2249.1999.00957.x
Zheng L Fisher G Miller RE Peschon J Lynch DH Lenardo MJ Induction of apoptosis in mature T cells by tumor necrosis factor Nature 1995 377 348 351 7566090 10.1038/377348a0
Fagiola U Cossarizza A Scala E Fanales-Belasio E Ortolani C Cozzi E Monti D Franceschi C Paganelli R Increased cytokine production in mononuclear cells of healthy elderly people Eur J Immunol 1993 23 2375 2378 8370415
Brunnsgaard H Andersen-Ranberg K Hjelmborg JB Pedersen BK Jeu B Elevated tumor necrosis factor alpha and mortality in centenarians Amer J Med 2003 115 278 283 12967692 10.1016/S0002-9343(03)00329-2
Trzonkowski P Myslizska J Godlewska B Szmit E Lukaszuk K Wieckiewicz J Brydak L Machala M Landowski J Mysliwski A Immune consequences of the spontaneous pro-inflammatory status in depressed elderly patients Brain Behav Immun 2004 18 135 148 14759591 10.1016/S0889-1591(03)00111-9
Njemini R Demanet C Mets T Inflammatory status as an important determinant of heat shock protein 70 serum concentration during aging Biogerontology 2004 5 31 38 15138379 10.1023/B:BGEN.0000017684.15626.29
Penninx BWJH Kritchevsky SB Newman AB Nicklas BJ Simonsick EM Rubin S Nevitt M Visser M Harris T Pahor M Inflammatory markers and incident mortality limitation in the elderly J Amer Gerontol Soc 2004 52 1105 1113 10.1111/j.1532-5415.2004.52308.x
De Maat MPM Bladbjerg EM Hjelmborg JVB Bathum L Jespersen J Christensen K Genetic influence on inflammation variables in the elderly Atheroscler Thromb Vasc Biol 2004 24 2168 2173 10.1161/01.ATV.0000143856.01669.e7
Kohut ML Senchina DS Madden KS Martin AE Felten DL Moynihan JA Age effect on macrophage function vary by tissue site, nature of stimulant, and exercise behavior Exp Gerontol 2004 39 1347 1360 15489058 10.1016/j.exger.2004.07.001
Savioli S Capri M Scarcella E Mangherini S Franca I Volterra V De Ronchi D Marini M Bonafe M Franceschi C Monti D Age-dependent changes in the susceptibility to apoptosis of peripheral blood CD4+ and CD8+ T lymphocytes with virgin or memory phenotype Mech Ageing Dev 2003 124 409 418 12714247 10.1016/S0047-6374(03)00016-2
Aggarwal S Gollapudi S Yel L Gupta AS Gupta S TNF-α-induced apoptosis in neonatal lymphocytes: TNFRp55 expression and downstream pathways of apoptosis Genes Immunity 2000 1 271 279 11196704 10.1038/sj.gene.6363674
Gupta S Kim C Yel L Gollapudi S A role of Fas-associated death domain (FADD) in increased apoptosis in aged humans J Clin Immunol 2004 24 24 29 14997030 10.1023/B:JOCI.0000018059.56924.99
Gupta S Bi R Kim C Yel L Chiplunkar S Gollapudi S A role of NF-κB signaling pathway in increased tumor necrosis factor-α-induced apoptosis of lymphocytes in aged humans Cell Death Diff 2005 12 177 183 10.1038/sj.cdd.4401557
Gupta S A role of inhibitor of apoptosis (IAP) proteins in increased lymphocyte apoptosis in aged humans Mech Ageing Dev 2004 125 99 101 15037009 10.1016/j.mad.2003.11.001
Whisler RL Beiqing L Chen M Age-related decreases in IL-2 production by human T cells are associated with impaired activation of nuclear transcriptional factors AP-1 and NF-AT Cell Immunol 1996 169 185 195 8620546 10.1006/cimm.1996.0109
Pahlavani M Harris MD The age-related changes in DNA binding activity of AP-1, NF-κB, and Oct-1 transcription factors in lymphocytes from rats Age 1996 19 45 54
Trebilcock GU Ponnappan U Evidence for lowered induction of nuclear factor kappa B in activated human T lymphocytes during aging Gerontology 42 137 146 8796372
Ponnappan U Zhong M Trebilcock GU Decreased proteosome-mediated degradation in T cells from the elderly: A role in immune senescence Cell Immunol 1999 192 167 174 10087185 10.1006/cimm.1998.1418
Sallusto F Geginat J Lanzavecchia A Central memory and effector memory T cell subsets: Function, generation, and maintenance Ann Rev Immunol 2004 22 745 763 15032595 10.1146/annurev.immunol.22.012703.104702
Kaech SM Ahmed R Memory CD8+ T cell differentiation: initial antigen encounter triggers a developmental program in naïve cells Nature Immunol 2001 2 415 422 11323695
Moser B Loetscher P Lymphocyte traffic control by chemokines Nature Immunol 2001 2 123 128 11175804 10.1038/84219
Schluns KS Lefrancois L Cytokine control of memory T-cell development and survival Nature Rev Immunol 2003 3 269 279 12669018 10.1038/nri1052
Kataoka T Budd RC Holler N Thome M Martinon F Irmler M Burns K Masopust D Vezys V Marzo AL Lanzavecchia A Preferential localization of effector memory cells in nonlymphoid tissue Science 2001 291 2413 2417 11264538 10.1126/science.1058867
Arbones SL Ord DC Ley K Ratech H Maynard-Curry C Otten G Capon DJ Tedder TF Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficienct mice Immunity 1994 1 247 260 7534203 10.1016/1074-7613(94)90076-0
Campbell JJ Bowman EP Murphy K Youngman KR Siani MA Thompson DA Wu L Zlotnik A Butcher EC 6-C-kine (SLC), a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is an agonist for MIP-13β receptor (CCR7) J Biol Chem 1998 414 1053 1059
Sallusto F Lenig D Forster R Lipp M Lanzavecchia A Two subsets of memory T lymphocytes with distinct homing potentials and effector functions Nature 1999 401 708 712 10537110 10.1038/44385
Monteiro J Baltiwala F Ostere H Gregersen PK Shortened telomere in clonally expanded CD28-CD8+ T cells imply a replicative history that is distinct from there CD28+CD8+ counterparts J Immunol 1996 162 6572 6579
Weninger W Crowley MA Manjunath N von Andriane UH Migratory properties of naïve, effector, and memory CD8 (+) T cells J Exp Med 2001 194 953 966 11581317 10.1084/jem.194.7.953
Gupta S Bi R Su K Yel L Chiplunkar S Gollapudi S Characterization of naïve, memory, and effector CD8+ T cells: Effect of age Exp Gerontol 2004 39 545 550 15050289 10.1016/j.exger.2003.08.013
Gupta S Su H Bi R Gollapudi S Differential sensitivity of naïve and memory subsets of human CD8+ T cells to TNF-α-induced apoptosis
Gupta S Gollapudi S Molecular mechanisms of TNF-α-induced apoptosis in aging human T cell subsets Int J Biochem Cell Biol 2004 37 1034 1042 15743676 10.1016/j.biocel.2004.08.013
Brzezinska A Magalska A Szybinska A Sikora E Proliferation and apoptosis of human CD8+CD28+ and CD8+CD28- lymphocytes during aging Exp Gerontol 2004 39 539 544 15050288 10.1016/j.exger.2003.09.026
Posnett DN Sinha R Kabak S Russo C Clonal populations of T cells in normal elderly humans: The cell equivalent to "benign monoclonal gammopathy" J Exp Med 1994 179 609 618 8294871 10.1084/jem.179.2.609
Effros RB Boucher N Porter V Zhu X Spaulding C Walford RL Kronenberg M Cohen D Schachter F Decline in CD28+ T cells in centenarians and in long-term T cell cultures: A possible cause of both in vivo and in vitro immunosenescence Exp Gerontol 1994 29 601 609 9435913 10.1016/0531-5565(94)90073-6
Monteiro J Baltiwala F Ostrer H Gregersen PK Shortened telomeres in clonally expanded CD28-CD8+ T cells imply a replicative history that is distinct from their CD28+CD8+ counterparts J Immunol 1996 156 3587 3590 8621891
Bandres E Merino J Vazquez S Inoges S Moreno C Subira ML Sanchez-Ibarrola A The increase of IFN-γ production through aging correlates with the expanded CD8+CD28-CD57+ subpopulation Clin Immunol 2000 96 230 235 10964541 10.1006/clim.2000.4894
Nociari MM Telford W Russo C Postthymic development of CD28-CD8+ T cell subset: age-associated expansion and shift from memory to naïve phenotype J Immunol 1999 162 3327 3335 10092786
Saurwein-Teissl M Lung TL Marx F Gschosser C Asch E Blasko I Parson W Bock G Schonitzer D Trannoy E Grubeck-Loebenstein B Lack of antibody production following immunization in old age:Association with CD8+CD28- T cell clonal expansions and an imbalance in the production of Th1 and Th2 cytokines J Immunol 2002 168 5893 5899 12023394
Herndon FJ Hsu HC Mountz JD Increased apoptosis of CD45RO- T cells with aging Mech Ageing Dev 1997 94 123 134 9147365 10.1016/S0047-6374(97)01882-4
Miyawaki T Uehara T Nabu R Tsuji T Yachie A Yonehara Y Taniguchi N Differential expression of apoptosis-related Fas antigen on lymphocyte subpopulations in human peripheral blood J Immunol 1992 149 3753 3758 1385530
Shinohara S Sawada T Nishioka Y Tohma S Kisaki T Inou T Ando K Ikeda M Fuji H Ito K Differential expression of Fas and Bcl-2 protein on CD+ T cells, CD8+ T cells and monocytes Cell Immunol 163 303 308 7541728 10.1006/cimm.1995.1130
Iwai K Miyawaki T Takizawa T Kondo A Ohta K Yachi A Seki H Taniguchi N Differential expression of bcl-2 and susceptibility to anti-Fas-mediated death in peripheral blood lymphocytes, monocytes and neutrophils Blood 1994 84 1201 1208 7519477
Yoshino K Kondo E Cao L Takahashi K Hayashi K Nomura S Akagi T Inverse expression of Bcl-2 protein and Fas antigen in lymphoblasts in peripheral nodes and activated peripheral blood T and B lymphocytes 1994 83 1856 1861 7511441
Liu Y-J Oldsfield S MacLennan IC Memory B cells in T cell-dependent antibody responses colonize the splenic marginal zones Eur J Immunol 1988 18 355 362 3258564
Smith KG Hewitson TD Nossal GJ Tarlinton DM The phenotype and fate of the antibody-forming cells of the splenic foci Eur J Immunol 1996 24 444 448 8617316
Liu Y-J Banchereau J The paths and molecular control of peripheral B cell development Immunologist 1996 4 55 60
Liu Y-J Barthelemy C de Boutetler O Arpin C Durand I Banchereau J Memory B cells from human tonsils colonize mucosal epithelium and directly present antigen to T cells by rapid upregulation of B7-1 and B7-2 Immunity 1995 2 239 248 7535180 10.1016/1074-7613(95)90048-9
Agematsu K Hokibara S Nagumo H Komiyama A CD27: a memory B cell marker Immunology Today 2000 21 204 206 10782048 10.1016/S0167-5699(00)01605-4
Ghia P Melchers F Rolink AG Age-dependent changes in B lymphocyte development in man and mouse Exp Gerontol 2000 35 159 65 10767576 10.1016/S0531-5565(99)00095-9
Banerjee M Mehr R Belelovsky A Spencer J Walters D Age-and tissue-specific differences in human germinal center B cell selection revealed by analysis of IgVH gene hypermutation and linage tree Eur J Immunol 2000 20 1947 1957
|
16115325
|
PMC1236953
|
CC BY
|
2021-01-04 16:36:31
|
no
|
Immun Ageing. 2005 Aug 23; 2:12
|
utf-8
|
Immun Ageing
| 2,005 |
10.1186/1742-4933-2-12
|
oa_comm
|
==== Front
Int Semin Surg OncolInternational seminars in surgical oncology : ISSO1477-7800BioMed Central London 1477-7800-2-191616229310.1186/1477-7800-2-19Case ReportSynchronous early primary adenocarcinoma of both rectum and gallbladder. Report of a case Sakellaridis Timothy [email protected] Stavros [email protected] Christos [email protected] Surgical Department, 401 General Military Hospital of Athens, Athens, Greece2005 14 9 2005 2 19 19 4 8 2005 14 9 2005 Copyright © 2005 Timothy et al; licensee BioMed Central Ltd.2005Timothy 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
Synchronous early primary cancers are rare and in addition synchronous adenocarcinoma of both rectum and gallbladder is extremely rare.
Case report
We report an unusual case of synchronous early primary adenocarcinoma of rectum and gallbladder. The patient was a 72-year-old woman with complaints of bloody stools and constipation. An endoscopy revealed adenocarcinoma of the lower rectum. A through preoperative investigation showed also cholelithiasis. The patient underwent abdominoperineal resection and cholecystectomy. The histopathological diagnosis was well to middle differentiate adenocarcinoma of the gallbladder (T2, N0, M0; stage II) and middle differentiate adenocarcinoma of the rectum (T2, N0, M0; stage II).
Conclusion
For the cases of extracolonic primary cancer associated with colorectal primary carcinoma, Warren and Gates' diagnostic criteria are used. All patients with colorectal carcinoma, should undergo a throughout preoperative examination to exclude the possibility of synchronous early primary cancers.
Adenocarcinoma of gallbladdercolorectal adenocarcinomasynchronous primary tumor
==== Body
Introduction
Synchronous early primary cancers are rare. In recent years multiple primary tumors have been documented more frequently. A review of autopsies of 659 cases of multiple primary tumors showed the colon to be the organ most frequently involved, especially among the aged [1,2].
Colorectal cancer is the third most common cancer in males and females. The cumulative lifetime risk of developing colorectal cancer is about 6%. However, it still accounts for 11% of cancer deaths. The risk of colorectal cancer increases with age. Although carcinoma of the gallbladder is a rare tumor, it is the most common malignancy of the biliary system and the fifth most common cancer of the gastrointestinal tract mostly among patients in their seventh and eighth decades.
We report a case of a female who presented with adenocarcinoma of rectum and cholelithiasis. The histopathology revealed early primary adenocarcinoma of both the rectum and gallbladder.
Case report
A 72-year-old Caucasian woman was complaining of bloody stools and constipation for three months. There was a palpable mass 3–4 cm from the anal verge. A colonoscopy showed a tumor 3 cm from the anal margin with no other indication of multiple synchronous tumors in the colon. Biopsies of the tumor were positive for adenocarcinoma. An investigation with upper gastrointestinal endoscopy, computed tomography of chest and abdomen showed cholelithiasis with thickening of gallbladder anterior wall (arrow in the figure 1). No evidence of metastatic disease documented. Blood examinations showed anemia (Ht: 28,5% and Hb: 9,5 g/dL) and the rest laboratory evaluation was normal. The carcinoembryonic antigen (CEA) was: 15,9 ng/ml (range 0,0–10,0). She had a cholecystectomy and abdominoperineal resection of the rectum. The histopathological diagnosis was a moderately differentiated adenocarcinoma of the gallbladder (T2, N0, M0; stage II) and differentiated moderately adenocarcinoma of the rectum (pT2, N0, M0 in TNM/UICC, stage A of the Dukes staging system and stage B1 of the modified Astler – Coller classification). No adjuvant chemotherapy was required post-operatively. The patient joined a five-year follow up programme and is doing well six months after surgery.
Discussion
Synchronous adenocarcinoma of both rectum and gallbladder is extremely rare. Review of the pertinent literature revealed no more than five cases worldwide [3-7]. Synchronous cancers are defined as those diagnosed at the same time or within six months; cancers are considered metachronous when the second tumor is diagnosed more than six months after the first1. For the cases of extracolonic primary cancer associated with colorectal primary carcinoma, Warren and Gates' diagnostic criteria are used [1]: 1) Each of the tumors have to present a definite picture of malignancy; 2) Each have to be distinct; and 3) The probability that one is metastatic from the other has to be excluded.
All patients with colorectal carcinoma, in addition to the history and physical examination, should undergo a throughout preoperative examination, which should include chest radiograph, complete blood cell count, liver function tests, electrolytes, urinalysis, carcinoembryonic antigen (CEA) determination, fecal occult blood testing (FOBT) (using Hemoccult or Hemoquant), full colonoscopy, double contrast barium enemas, endorectal ultrasound for rectal tumors and abdominopelvic computed tomography scan. In colorectal cancer patients without evidence of distant metastases, a complete meticulous surgical exploration should not be denied. When multiple lesions or evidence of distant metastases are found, the operation should be individualized according to the location, state of spread and the patient's condition. Postoperative follow-up should include a complete blood tests and throughout examination of the remainder colon at least every 6 months. Metachronous cancer may appear as long as 15 years after the first cancer has been removed [1]. Extracolonic primary cancer is reported more frequently in the skin, stomach, breast, urinary bladder and prostate, and it may occur as long as 17 years before, or 20 years after, the diagnosis of the colorectal cancer [1]. Those findings illustrate the pitfalls in assuming any lesion to be a metastasis or a recurrence without pathologic confirmation.
The prognosis of these patients appears to equal or to be only slightly worse than, that for a single colorectal cancer [1]. There have been evidence that in some cases this entity can be ascribed to a genetic defect or an unknown carcinogenic agent or be part of cancer family syndrome [4,8,9].
Figure 1 Computed tomography, showing cholelithiasis with thickening of gallbladder anterior wall (arrow in the figure).
==== Refs
Lee TK Barringer M Myers RT Sterchi JM Multiple primary carcinomas of the colon and associated extracolonic primary malignant tumors Ann Surg 1982 195 501 7 7065754
Arai T Sawabe M Takubo K Kanazawa K Esaki Y Multiple colorectal cancers in the elderly: a retrospective study of both surgical and autopsy cases Gastroenterol 2001 36 748 52 10.1007/s005350170016
Kondo T Ishii Y Kobayashi S Moriwaki M Sakurai H Okamura H A case of triple simultaneous cancer of the esophagus, gallbladder and colon Gan No Rinsho 1985 31 1849 53 4087393
Schmid KW Glaser K Wykypiel H Feichtinger H Synchronous adenocarcinoma of the transverse colon, the gallbladder and the vermiform appendix Klin Wochenschr 66 1093 6 1988 Nov 1 3236758 10.1007/BF01711925
Tamura M Shinagawa M Funaki Y Synchronous triple early cancers occurring in the stomach, colon and gallbladder Asian J Surg 2003 26 46 9 12527496
Tomin R Simultaneous cancer of the rectum and gallbladder Srp Arh Celok Lek 1965 93 323 7 5861499
Yazawa K Tsuge Y Takahashi C Ichijo M Matsuzaki O A case of synchronous triple early cancer occurring in the stomach, colon and gallbladder Nippon Shokakibyo Gakkai Zasshi 1997 94 440 4 9216227
Mosca F Stracqualursi A Lipari G Persi A Latteri S Multiple primary malignant neoplasms. Report of a rare case with 5 metachronous tumors Chir Ital 2001 53 133 9 11280822
Tokunaga A Onda M Shimizu Y Yoshiyuki T Nishi K Matsukura N Non-simultaneous primary cancers in five different organs in a case of family syndrome Jpn J Cancer Res 1987 78 242 50 3106280
|
16162293
|
PMC1236954
|
CC BY
|
2021-01-04 16:38:37
|
no
|
Int Semin Surg Oncol. 2005 Sep 14; 2:19
|
utf-8
|
Int Semin Surg Oncol
| 2,005 |
10.1186/1477-7800-2-19
|
oa_comm
|
==== Front
J Neuroengineering RehabilJournal of NeuroEngineering and Rehabilitation1743-0003BioMed Central London 1743-0003-2-271612488010.1186/1743-0003-2-27ResearchShedding light on walking in the dark: the effects of reduced lighting on the gait of older adults with a higher-level gait disorder and controls Kesler Anat [email protected] Gregory [email protected] Talia [email protected] Leor [email protected] Nir [email protected] Jeffrey M [email protected] Movement Disorders Unit, Department of Neurology, Tel-Aviv Sourasky Medical Center, Tel-Aviv, Israel2 Department of Ophthalmology, Tel-Aviv Sourasky Medical Center, Tel-Aviv, Israel3 Department of Physical Therapy, Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel4 Department of Neurology, Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel5 Division on Aging, Harvard Medical School, Boston, MA, USA2005 28 8 2005 2 27 27 5 4 2005 28 8 2005 Copyright © 2005 Kesler et al; licensee BioMed Central Ltd.2005Kesler 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 study the effects of reduced lighting on the gait of older adults with a high level gait disorder (HLGD) and to compare their response to that of healthy elderly controls.
Methods
22 patients with a HLGD and 20 age-matched healthy controls were studied under usual lighting conditions (1000 lumens) and in near darkness (5 lumens). Gait speed and gait dynamics were measured under both conditions. Cognitive function, co-morbidities, depressive symptoms, and vision were also evaluated.
Results
Under usual lighting conditions, patients walked more slowly, with reduced swing times, and increased stride-to-stride variability, compared to controls. When walking under near darkness conditions, both groups slowed their gait. All other measures of gait were not affected by lighting in the controls. In contrast, patients further reduced their swing times and increased their stride-to-stride variability, both stride time variability and swing time variability. The unique response of the patients was not explained by vision, mental status, co-morbidities, or the values of walking under usual lighting conditions.
Conclusion
Walking with reduced lighting does not affect the gait of healthy elderly subjects, except for a reduction in speed. On the other hand, the gait of older adults with a HLGD becomes more variable and unsteady when they walk in near darkness, despite adapting a slow and cautious gait. Further work is needed to identify the causes of the maladaptive response among patients with a HLGD and the potential connection between this behavior and the increased fall risk observed in these patients.
gaitvariabilityvisionfall riskaginglightingHigher-Level Gait Disorders
==== Body
Introduction
Many older adults have an impaired gait that does not appear to be a result of any well defined disease [1]. In their review of patients attending a neurology clinic, Sudarsky et al. reported that the cause of the gait disturbance was unknown, even after neuro-imaging, in about 10–20 percent of older adults with a disturbed gait [2,3]. In a study of the "oldest old" (age range 87 to 97 years) in the Netherlands, Bloem et al. observed that about 20 percent of those studied had a normal gait, 69 percent had a gait disorder due to known disease, and about 11 percent of the subjects had an idiopathic "senile gait disorder", i.e., a gait disorder of unknown origin [4]. Of note, those subjects with a gait disorder of unknown origin had a higher risk of mortality during a five year follow up period, compared to the group of age-matched subjects who had a normal gait [5], suggesting that the origin of this gait disorder is not benign.
Nutt et al. coined the term "higher-level" gait disorders (HLGD) to refer to an altered gait that is not a result of lower extremity or peripheral dysfunction and cannot be attributed to well defined chronic disease [6,7]. One common example of a HLGD is the idiopathic "cautious" gait of the elderly or the "senile gait" disorder [6,7]. A "cautious" gait is typically marked by mild to moderate slowing, reduced stride length, and mild widening of the base of support [7]. Previous studies have shown that older adults with a cautious gait and HLGD also walk with increased stride variability and unsteadiness, have an excessive fear of falling that appears to be related to this increased stride variability, and have an increased risk of falls [8,9]. Further, the extrapyramidal, limbic systems, and the frontal lobe apparently play an important role, to different degrees, in what can be viewed as a multi-system neurodegenerative syndrome clearly different from "aging" [8]. Indeed, a three year prospective study found that gait and function deteriorated to a much greater extent among older adults with a HLGD, compared to controls, supporting the idea that this is a progressive neurodegenerative disorder [10]. However, the origin of the cautious gait in older adults with a HLGD remains largely unknown.
The gait changes observed in patients with a HLGD have features common to subjects who walk in the dark or with impaired vision, i.e., to others who might be walking cautiously. When vision is altered or lighting is reduced, subjects typically adapt a slower gait [11-13]. Variability of foot placement, at least during gait termination, may also be increased when lighting is not adequate [12]. These changes are reminiscent of the walking pattern of older adults with a HLGD and cautious gait. An elevated risk of falling has been associated with visual impairments, a problem that increases with age [14-17] and fall risk has also been associated with inadequate lighting, but the effects of vision and lighting have not been studied in older adults with a HLGD. To more fully characterize the gait of older adults with a HLGD and their reliance on visual input, we examined the effect of lighting changes on the gait of older adults with a HLGD and compared their response to that of healthy elderly controls. More specifically, we hypothesized that the response to near darkness may exacerbate gait instability and fall risk markers in these patients.
Methods
Participants
Twenty-two older adults between the ages of 70 and 90 years old who met previously established criteria for a HLGD [8,9] were included in the present study. Patients were recruited from among those who came to the Geriatric Outpatient Clinic or the Movement Disorders Unit at the Tel Aviv Sourasky Medical Center for evaluation of walking difficulties of unknown origin. All patients were mobile and walked independently at the time of assessment and all underwent a thorough general and neurological examination to ensure that subjects met the criteria of HLGD.
Patients were excluded if the cause of their gait disturbance could be readily established. Thus, patients with a history of clinically established stroke, Parkinson's disease, Alzheimer's disease, possible normal pressure hydrocephalus or other diagnosed neurodegenerative disorder, and patients with rest tremor or pronounced bradykinesia were excluded. Patients who were taking anti-parkinsonian or anti-spastic medications, or had orthostatic hypotension were also excluded. We also excluded patients with significant visual, peripheral, or vestibular disturbances, as well as patients with significant orthopedic disturbances. Patients with dementia according to the DSM IV criteria [18], history of psychiatric disease, or past use of dopamine receptor blocking agents (anti-psychotic medications) were excluded as well. In addition, we excluded patients with a history of traumatic head injury and/or loss of consciousness. In brief, no specific disorder could be diagnosed as the cause of the patients' complaint about his or her walking difficulties.
The patient population was compared to a group of twenty healthy controls of similar age. Control subjects were recruited from the community and from nearby elderly housing facilities or were spouses of outpatients. Subjects were included if they were between 70 and 90 years of age, reported normal walking function, had no obvious clinical impairment, and did not have significant cognitive impairment (Mini Mental State Examination >25 [19]). Subjects were excluded if they had any neurological disorder or any significant clinical history likely to affect their gait (e.g., stroke).
The study was approved by the Human Studies committee of the Tel-Aviv Sourasky Medical Center. All subjects provided informed written consent according to the declaration of Helsinki prior to entering the study.
Subject Characteristics and Assessment of Vision
The Mini Mental State Examination (MMSE) [19] and the Geriatric Depression Scale (GDS) [20] were administered to probe the mental health of the subjects. Body-mass-index (BMI) was determined and Charlson's co-morbidity index was used to quantify general health status; scores closer to zero reflect better health [21].
Three aspects of vision were evaluated: 1) visual acuity, using the Snellen vision chart, 2) color blindness, using Ishihara pseudochromatic color test [22], and 3) contrast sensitivity. Previous studies have indicated that visual acuity and contrast sensitivity, a robust indicator of functional vision [23], are associated with an increased risk of falls among the elderly [16,24-26]. Visual acuity scores were stratified in normal (i.e., good or mild decline, 6/6–6/12) and abnormal (>6/15). Contrast sensitivity was measured using a wall mounted clinical chart, a standard clinical tool (Vistech VCTS 6000). The chart contains 5 rows of 9 printed circular patches each of which displays a sine wave grating. There are 5 spatial frequencies across the 5 rows (1.5, 3, 6, 12, and 18 cycles per degree). The chart luminance was standardized according to the light meter supplied with the chart. The last patch on which the patient correctly identified the direction of the gratings was recorded for each frequency. For all tests, each eye was examined separately. The function of the better eye was used in all analysis, since both eyes were used during walking. The eye examinations were performed by a neuro-ophthalmologist who was blinded to the gait measures in a subset of subjects who were selected at random.
Walking Protocol
Subjects were instructed to walk on level ground under usual lighting conditions (1000 lumens) and in near darkness (5 lumens). Although there are advantages to performing these tests in a random order, the usual lighting condition was always performed first. If anything, this should maximize safety and minimize the effects of the walking in darkness; since this condition is always performed second, subjects have more time to adapt to the environment and walking conditions. Between the two walks, subjects sat and rested for at least two minutes. In order to control the lighting conditions, testing took place in a large, quiet, empty room. The straight walking path was 9 meters long. Subjects walked along the path six times and were told to turn around and continue walking when they reached the end of the path. All subjects were tested in the same environment. Subjects were "guarded" by a research assistant who walked a few steps away from the subject, making sure not to interfere or set the pace. Study subjects were not aware of the specific questions of this investigation.
Assessment of Gait Dynamics
Previously described methods were used to quantify gait variability and evaluate gait dynamics of each walk [9,27,28]. Briefly, to measure the gait rhythm and the timing of the gait cycle (i.e., the stride time and the swing time), a computerized force-sensitive system was used to evaluate gait and stride-to-stride variability [29,30]. The system measures the forces underneath the foot as a function of time. The system consists of a pair of shoes and a recording unit. Each shoe contains 8 load sensors that cover the surface of the sole and measure the vertical forces under the foot. The recording unit (19 × 14 × 4.5 cm; 1.5 kg) is carried on the waist. Plantar pressures under each foot are recorded at a rate of 100 Hz. Measurements are stored in a memory card during the walk and, after the walk, are transferred to a personal computer for further analysis. Subsequently, the digitized data were transferred to a computer workstation for analysis using software that extracts the initial and end contact time of each stride and determines stride and swing times. To focus on the assessment of the dynamics of continuous, "normal" walking and each subject's "intrinsic" dynamics and to ensure that the analysis was not influenced by atypical strides (e.g., the turning at the end of the room), a median filter was applied to each subject's time series to remove data points that were three standard deviations greater than or less than the median value [31]. Subsequently, the average stride time, average swing time, stride time variability, and swing time variability were determined. Variability was calculated using the coefficient of variation (CV) of each subject's stride time or swing time, e.g., (100 × standard deviation of stride time)/(mean stride time). Stride-to-stride variability reflects gait unsteadiness and arrhythmicity and has been shown to prospectively predict falls [31-34]. The time to walk the 54 meters, the walk time, was also measured. The values for the left and right feet were highly correlated; for brevity, we report values from only right foot.
Statistical Analysis
Descriptive statistics are reported as mean ± SD or %. We used the Student's t and Chi-square tests to compare the patient and control subjects with respect to different background characteristics (e.g., age, gender, vision) and Spearman's correlation coefficient to quantify correlations among measures. To evaluate the effect of lighting on gait parameters and to compare the groups, we used Mixed Effects Models for repeated measures. For each gait parameter, a separate model was applied. The dependent variable was the gait parameter and the independent variables were the group (patients or controls), the walking condition (i.e., light or near darkness), and the interaction term group × lighting condition. A p-value ≤ 0.05 (two-sided) was considered statistically significant. All statistical analyses were performed using SPSS 11.5 and SAS 8.2 (Proc Mixed).
Results
Table 1 summarizes the general characteristics of the two study groups. Patients and controls were similar with respect to age, gender, body-mass-index, and depressive symptoms. MMSE scores were slightly, but significantly lower in the patients, but all subjects from both groups scored a 26 or higher (i.e., they were non-demented). Charlson scores were higher in the patients, but the scores were generally low and close to 0 (low co-morbidity) in both groups. As shown in Table 2, measures of visual acuity, color blindness and contrast sensitivity were similar in the patients and the controls.
Table 1 Subjects characteristics*
Patients with HLGD (n = 22) Controls (n = 20)
Age (yrs) 80.7 ± 4.1 80.6 ± 6.3
Gender (% male) 73% 65%
Body-mass-index (kg/m2) 26.6 ± 4.9 25.1 ± 2.9
Mini Mental State Exam (MMSE) 28.1 ± 1.3 29.4 ± 0.9
Geriatric Depression Scale 5.6 ± 4.7 3.8 ± 2.6
Charlson Comorbidity Score 0.0 ± 0.0 0.5 ± 0.7
*Subject characteristics were not different in the two groups (p > 0.13), except that the MMSE and the Charlson score tended to be slightly different in the patients (p < 0.01). Values are mean ± SD or %, as indicated. HLGD: Higher-level gait disorder.
Table 2 Measures of vision in the two study groups*
Patients with HLGD (n = 11) Controls (n = 15)
Visual Acuity (% normal) 82% 73%
Color Vision Test 8.2 ± 2.5 8.5 ± 1.3
Contrast Sensitivity Test: low spatial frequency 4.8 ± 0.7 4.5 ± 0.7
Contrast Sensitivity Test: intermediate spatial frequency 4.4 ± 0.7 4.2 ± 1.2
Contrast Sensitivity Test: high spatial frequency 1.8 ± 1.5 1.7 ± 1.5
*p > 0.19 for all comparisons.
Under normal lighting conditions, HLGD patients took more time to complete the walk and walked with an increased stride time, reduced swing time, and increased stride-to-stride variability of the stride and swing time, compared to the control subjects (p < 0.01) (see Table 3). Compared to normal lighting conditions, both patients and controls required significantly more time to complete the walk when walking in near darkness (p < 0.005). Walk times increased by 14.3% in the controls and by 15.8% in the patients, in other words, by similar amounts (p = 0.828). Among the control subjects, walking in near darkness did not significantly affect the average stride time, the average swing time, or the stride-to-stride variability of these measures (p > 0.29).
Table 3 Effects of lighting on gait
Patients with HLGD (n = 22) Controls (n = 20)
Normal Lighting Near Dark (P-value*) Normal Lighting Near Dark (P-value*)
Average Stride Time (sec) 1.30 ± 0.17 1.29 ± 0.15 (0.376) 1.17 ± 0.12 1.17 ± 0.12 (0.912)
Stride Time Variability (%) 5.6 ± 2.3 6.8 ± 2.3 (0.005) 3.6 ± 1.2 4.1 ± 1.9 (0.295)
Average Swing Time (%) 33.9 ± 2.7 32.5 ± 3.7 (<0.001) 35.7 ± 3.0 35.5 ± 3.2 (0.015)
Swing Time Variability (%) 7.0 ± 2.9 10.1 ± 4.7 (<0.001) 4.5 ± 2.4 5.1 ± 2.5 (0.365)
Walk Time (sec) 106.3 ± 44.2 124.4 ± 57.1 (<0.001) 72.8 ± 20.8 84.2 ± 33.5 (0.013)
*P-values shown in parentheses are based on within group comparisons between near dark and normal lighting. All measures of gait were different (p < 0.01) in the two subject groups, both under normal lighting and in near darkness. Walk time is the time to complete the 54 meter walk.
In contrast to the control group, the gait of the patients with a HLGD became more abnormal when they walked in near darkness. There was no change in the average stride time when the patients walked in near darkness (p = 0.376), but the average swing time significantly decreased (p < 0.001), stride time variability increased (p = 0.005) and swing time variability increased (p < 0.001). As might be expected from Table 2, the group × lighting condition interaction term was significant for the average swing time (p = 0.015) and swing time variability (p = 0.005), indicating that near darkness affected the patients more than the controls (see Figure 1), despite similar increases in walk time. As noted above, all of the group differences in gait observed during normal lighting conditions persisted during walking in near darkness and the gap between the two groups widened.
Figure 1 Effects of near darkness on stride time, stride time variability, and swing time variability in the two groups. For both groups, the average stride time was not affected by the change in lighting (p > 0.37). During walking in near darkness, variability measures were not significantly changed in the healthy controls (p > 0.29), but in the patients, stride time variability (p = 0.005) and swing time variability (p < 0.001) became significantly larger, compared to the values measured normal lighting.
Among the background and vision measures evaluated (e.g., age, gender, visual acuity, contrast sensitivity), the MMSE and the Charlson scores were the only measures that were significantly different and could, therefore, potentially mediate the group-specific changes in gait during walking in near dark. The MMSE was not correlated with the change in any of the gait measures observed during near dark walking (p > 0.17). Similarly, the Charlson scores did not explain the change in any of the gait measures (p > 0.07).
Subjects who took longer to complete the walk with normal lighting generally showed a relative increase in the walk time during near dark walking (r = 0.48; p = 0.001), whereas the changes in stride time variability and swing time variability were not significantly associated with the baseline values of these measures (p > 0.17). The change in stride time variability and swing time variability were moderately correlated with each other (r = 0.38; p = 0.018). The change in swing time variability was moderately correlated with the change in walk time (r = -0.40; p = 0.016), but the change in stride time variability was not correlated with the change in walk time (p = 0.12).
Discussion
To summarize, key findings of this study are: 1) Both healthy older adults and older adults with a HLGD walk more slowly under diminished lighting conditions; 2) Diminished lighting does not increase the gait variability of healthy older adults; and 3) In patients with a HLGD, diminished lighting significantly increases gait variability. In the following paragraphs, we attempt to interpret these findings and discuss their implications for understanding HLGD, the role of vision in gait, and the relationship between visual impairment and increased fall risk in older adults.
Perhaps the simplest way to interpret the slower walking that occurs during diminished lighting conditions is that subjects are adapting a more cautious gait in response to the reduced lighting. This would be consistent with previous studies that suggest that gait slows down in the absence of sufficient visual input [11-13], perhaps to increase safety. This response of the healthy older adults to walking in near darkness also parallels the effects of an attention demanding task on the gait of healthy young and older adults [35,36]. When healthy young or older adults are asked to walk and perform an additional task simultaneously, gait speed is reduced, but there is no effect on gait variability [35,36]. From this perspective, one could suggest that walking in near darkness requires greater attention than walking under normal lighting.
When older adults with neurodegenerative disease or those with an increased risk of falls walk while simultaneously performing an attention demanding task, two things happen: 1) they slow down, like their healthy peers, and 2) stride variability increases [28,35-37]. These are the effects that were seen in the present study when the patients with a HLGD walked in near darkness. As noted, a possible explanation for this behavior, therefore, is that for the patients with a HLGD, walking in the dark is an attention demanding task. Alternatively, one could suggest that walking in near darkness reduces self-efficacy of walking in patients with HLGD, because they are already predisposed to fear of falling, but not in the controls, who do not have a marked concern about their gait. This could explain the disparate response of the two groups. Indeed, in older adults with a HLGD, fear of falling has been associated with stride time variability [9]. However, if this were the only contributing factor, one might have expected to see a larger reduction in walk times in the patient group, compared to the control, whereas the relative increases in walk times during near darkness were similar in the two groups.
Another potential explanation for the increased stride variability observed in the patients with a HLGD is based on the relationship between gait speed, stride length and stride frequency, on the one hand, and stride variability on the other [38-40]. At least in certain populations, some investigators suggest that variability of stride time and stride length becomes greater at slower walking speeds. One could argue that the increased variability observed in the patients with a HLGD in near darkness is simply a byproduct of their reduced walking speed. This explanation is, however, likely to be incomplete or incorrect. First, as noted, the healthy controls slowed down when walking in near darkness by virtually the same extent as that seen in the patient group, yet variability measures were unchanged in the controls. A reduced gait speed by itself does not necessarily lead to increased variability (as is the case for the response of healthy subjects to an attention demanding task, as discussed above). Second, in a study of healthy older adults and patients with Parkinson's disease, swing time variability was not affected by gait speed, even when gait speed was reduced by as much as 20% [30]. The increase in swing time variability observed among the patients with a HLGD during walking in near darkness cannot, therefore, be attributed to changes in gait speed.
Another way to view walking in near darkness is to consider it not simply as a test of vision, but as a challenge to other sensory feedback mechanisms that help to regulate gait. Walking may normally rely on visual, vestibular and proprioceptive feedback. When older adults or persons with deficits in balance are asked to close their eyes while standing on a balance platform, measures of sway and postural instability increase, both compared to eyes open conditions and compared to healthy young adults [41, 42]. These findings indicate that with aging, there is a greater reliance on visual feedback for maintenance of static balance; hence, when vision is removed, there is a large decrement in postural stability. We can apply similar reasoning to the present findings. In that case, one can interpret the results to suggest that in the patients with a HLGD, regulation of stride-to-stride variability relies on visual input and other feedback mechanisms are unable to fill in the gap that occurs when vision input is limited in near dark walking. This would suggest that patients with HLGD may have deficits in proprioception or vestibular function. Such deficits have not been identified in the present or previous studies of patients with HLGD [8,9]. It is possible, however, that these changes are relatively subtle and only surface when challenged.
The present study has several limitations. For example, we did not directly examine the affect of lighting on stress or fear of falling. Previous studies demonstrated that older patients with a HLGD have deficits in frontal lobe function, impairment in tests of balance and gait, and an increased risk of falls [8,9]. Future studies should assess if and how these factors contributed to the observed effects and how walking in darkness affects stress, anxiety and confidence in walking. It would also be helpful to evaluate other aspects of vision (e.g., peripheral vision) on a larger sample. We were not able to identify the specific factor that explained the increased sensitivity of the gait of patients with a HLGD to reduced lighting. Thus, the precise explanation for the further increase in stride-to-stride variability in near darkness in the patients with a HLGD remains to be determined.
Despite these limitations, the present findings shed light on the link between visual impairment, gait disturbances, and falls. Among older adults, falls are a major cause of morbidity and mortality [14]. Over one third of the adults aged 65 and over fall at least once each year [14] and among patients with a HLGD, falls are apparently much more frequent [9]. Previous studies have demonstrated that impaired vision is an important and independent risk factor for falls [14-17,24-26]. The present findings suggest a potential mechanism. With reduced vision or when walking in near darkness, perhaps two sides of the same coin, patients with an already increased risk of falls may further predispose themselves to falls and instability by increasing their stride-to-stride variability. A small perturbation could then take an already unstable system and cause a fall. Regardless of the precise explanation, the present results highlight the inappropriate response of patients with HLGD to reduced lighting conditions and suggest how this situation may aggravate gait instability and lead to falls in these older adults
Conflict of interest statement
The author(s) declare that they have no competing interests.
Contributors
A Kesler, G Leibovich, N Giladi, and JM Hausdorff designed the study. G Leibovich and T Herman participated in data collection. JM Hausdorff and L Gruendlinger helped with data analysis. A Kesler and JM Hausdorff drafted the manuscript. All authors helped with the interpretation of the results, reviewed the manuscript and participated in the editing of the final version of the manuscript.
Acknowledgements
We thank the participants for their time and effort and Dr. Lili Merdler for valuable assistance. This work was supported in part by grants from the NIA, NICHD and NCRR.
==== Refs
Whitman GT DiPatre PL Lopez IA Liu F Noori NE Vinters HV Baloh RW Neuropathology in older people with disequilibrium of unknown cause Neurology 1999 53 375 382 10430429
Sudarsky L Geriatrics: gait disorders in the elderly N Engl J Med 1990 322 1441 1446 2184358
Sudarsky L Gait disorders: prevalence, morbidity, and etiology Adv Neurol 2001 87 111 117 11347214
Bloem BR Gussekloo J Lagaay AM Remarque EJ Haan J Westendorp RG Idiopathic senile gait disorders are signs of subclinical disease J Am Geriatr Soc 2000 48 1098 1101 10983910
Nutt JG Marsden CD Thompson PD Human walking and higher-level gait disorders, particularly in the elderly Neurology 1993 43 268 279 8437689
Nutt JG Classification of gait and balance disorders Adv Neurol 2001 87 135 141 11347216
Giladi N Herman T Reider-Groswasser II Gurevich T Hausdorff JM Clinical characteristics of elderly patients with a cautious gait of unknown origin J Neurol 2005 252 300 306 15726273 10.1007/s00415-005-0641-2
Herman T Giladi N Gurevich T Hausdorff JM Gait instability and fractal dynamics of older adults with a "cautious" gait: why do certain older adults walk fearfully? Gait Posture 2005 21 178 185 15639397 10.1016/j.gaitpost.2004.01.014
Huber V Giladi N Herman T Peretz C Gruendlinger L Hausdorff JM A prospective study of older adults with a high level gait disorder: evidence for a neurodegenerative process Gait Posture 2005
Heasley K Buckley JG Scally A Twigg P Elliott DB Stepping up to a new level: effects of blurring vision in the elderly Invest Ophthalmol Vis Sci 2004 45 2122 2128 15223785 10.1167/iovs.03-1199
Perry SD Santos LC Patla AE Contribution of vision and cutaneous sensation to the control of centre of mass (COM) during gait termination Brain Res 2001 913 27 34 11532244 10.1016/S0006-8993(01)02748-2
Nakamura T Quantitative analysis of gait in the visually impaired Disabil Rehabil 1997 19 194 197 9184784
Guideline for the prevention of falls in older persons. American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopaedic Surgeons Panel on Falls Prevention J Am Geriatr Soc 2001 49 664 672 11380764 10.1046/j.1532-5415.2001.49115.x
Ivers RQ Norton R Cumming RG Butler M Campbell AJ Visual impairment and risk of hip fracture Am J Epidemiol 2000 152 633 639 11032158 10.1093/aje/152.7.633
Jack CI Smith T Neoh C Lye M McGalliard JN Prevalence of low vision in elderly patients admitted to an acute geriatric unit in Liverpool: elderly people who fall are more likely to have low vision Gerontology 1995 41 280 285 8537012
Seemongal-Dass RR James TE Atherley CE Guidelines for prevention of falls in people aged over 65. Guidelines should state that assessment of vision is important BMJ 2001 322 554 11263454 10.1136/bmj.322.7285.554
Association AP Diagnostic and statistical manual of mental disorders 1994 4 Washington, DC, American Psychiatric Association;
Folstein MF Folstein SE McHugh PR "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician J Psychiatr Res 1975 12 189 198 1202204 10.1016/0022-3956(75)90026-6
Yesavage JA Geriatric Depression Scale Psychopharmacol Bull 1988 24 709 711 3249773
Charlson ME Pompei P Ales KL MacKenzie CR A new method of classifying prognostic comorbidity in longitudinal studies: development and validation J Chron Dis 1987 40 373 383 3558716 10.1016/0021-9681(87)90171-8
Haskett MK Hovis JK Comparison of the Standard Pseudoisochromatic Plates to the Ishihara Color-Vision Test American Journal of Optometry and Physiological Optics 1987 64 211 216 3578487
Packer M Fine IH Hoffman RS Functional vision, contrast sensitivity, and optical aberrations Int Ophthalmol Clin 2003 43 1 3 12711898 10.1097/00004397-200343040-00003
Ivers R Cumming R Mitchell P Poor vision and risk of falls and fractures in older Australians: the Blue Mountains Eye Study N S W Public Health Bull 2002 13 8 10 12105666
Lord SR Clark RD Webster IW Physiological factors associated with falls in an elderly population J Am Geriatr Soc 1991 39 1194 1200 1960365
Lord SR Clark RD Webster IW Visual acuity and contrast sensitivity in relation to falls in an elderly population Age Ageing 1991 20 175 181 1853790
Schaafsma JD Giladi N Balash Y Bartels AL Gurevich T Hausdorff JM Gait dynamics in Parkinson's disease: relationship to Parkinsonian features, falls and response to levodopa J Neurol Sci 2003 212 47 53 12809998 10.1016/S0022-510X(03)00104-7
Sheridan PL Solomont J Kowall N Hausdorff JM Influence of executive function on locomotor function: divided attention increases gait variability in Alzheimer's disease J Am Geriatr Soc 2003 51 1633 1637 14687395 10.1046/j.1532-5415.2003.51516.x
Bazner H Oster M Daffertshofer M Hennerici M Assessment of gait in subcortical vascular encephalopathy by computerized analysis: a cross-sectional and longitudinal study J Neurol 2000 247 841 849 11151415 10.1007/s004150070070
Frenkel-Toledo S Giladi N Peretz C Herman T Gruedlinger L Hausdorff JM Treadmill walking as an external pacemaker to improve gait rhythm and stability in Parkinson's disease Mov Disord 2004
Hausdorff JM Rios D Edelberg HK Gait variability and fall risk in community-livingolder adults: a 1-year prospective study Arch Phys Med Rehabil 2001 82 1050 1056 11494184 10.1053/apmr.2001.24893
Hausdorff JM Edelberg HK Mitchell SL Goldberger AL Wei JY Increased gait unsteadiness in community-dwelling elderly fallers Arch Phys Med Rehabil 1997 78 278 283 9084350 10.1016/S0003-9993(97)90034-4
Maki BE Gait changes in older adults: predictors of falls or indicators of fear J Am Geriatr Soc 1997 45 313 320 9063277
Nakamura T Meguro K Sasaki H Relationship between falls and stride length variability in senile dementia of the Alzheimer type Gerontology 1996 42 108 113 9138973
Springer S Giladi N Simon ES Hausdorff JM Deficits in executive function in idiopathic elderly fallers: Association with fall risk Neurology 2004 62 A129 A129 10.1159/000080030
Yogev G Giladi N Gruendlinger L Baltadjieva R Simon ES Hausdorff JM Executive function, mental loading and gait variability in Parkinson's disease Journal of the American Geriatrics Society 2004 52 S3 S3 10.1111/j.1532-5415.2004.52003.x
Hausdorff JM Balash J Giladi N Effects of cognitive challenge on gait variability in patients with Parkinson's disease J Geriatr Psychiatry Neurol 2003 16 53 58 12641374
Danion F Varraine E Bonnard M Pailhous J Stride variability in human gait: the effect of stride frequency and stride length Gait Posture 2003 18 69 77 12855302 10.1016/S0966-6362(03)00030-4
Hausdorff JM Stride variability: beyond length and frequency Gait Posture 2004 20 304 15531178 10.1016/j.gaitpost.2003.08.002
Yamasaki M Sasaki T Torii M Sex difference in the pattern of lower limb movement during treadmill walking Eur J Appl Physiol Occup Physiol 1991 62 99 103 2022210 10.1007/BF00626763
Choy NL Brauer S Nitz J Changes in postural stability in women aged 20 to 80 years J Gerontol A Biol Sci Med Sci 2003 58 525 530 12807923
Gill J Allum JH Carpenter MG Held-Ziolkowska M Adkin AL Honegger F Pierchala K Trunk sway measures of postural stability during clinical balance tests: effects of age J Gerontol A Biol Sci Med Sci 2001 56 M438 M447 11445603
|
16124880
|
PMC1236955
|
CC BY
|
2021-01-04 16:37:40
|
no
|
J Neuroengineering Rehabil. 2005 Aug 28; 2:27
|
utf-8
|
J Neuroeng Rehabil
| 2,005 |
10.1186/1743-0003-2-27
|
oa_comm
|
==== Front
Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-241612239010.1186/1475-2859-4-24ReviewBiofilm reactors for industrial bioconversion processes: employing potential of enhanced reaction rates Qureshi Nasib [email protected] Bassam A [email protected] Thaddeus C [email protected] Patrick [email protected] Ian S [email protected] National Center for Agricultural Utilization Research, United States Department of Agriculture**, Agricultural Research Service, Fermentation Biotechnology Unit, 1815 N University Street, Peoria, IL 61604, USA2 Eastern Regional Research Center, United States Department of Agriculture, Agricultural Research Service, 600E Mermaid Lane, Wyndmoor, PA 19038, USA3 University of Illinois, Biotechnology & Bioengineering Group, Department of Food Science & Human Nutrition, 1207 W Gregory Drive, Urbana, IL 61801, USA4 Massey Univesity, Institute of Engineering & Technology, Palmerston North, New Zealand2005 25 8 2005 4 24 24 29 6 2005 25 8 2005 Copyright © 2005 Qureshi et al; licensee BioMed Central Ltd.2005Qureshi 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.
This article describes the use of biofilm reactors for the production of various chemicals by fermentation and wastewater treatment. Biofilm formation is a natural process where microbial cells attach to the support (adsorbent) or form flocs/aggregates (also called granules) without use of chemicals and form thick layers of cells known as "biofilms." As a result of biofilm formation, cell densities in the reactor increase and cell concentrations as high as 74 gL-1 can be achieved. The reactor configurations can be as simple as a batch reactor, continuous stirred tank reactor (CSTR), packed bed reactor (PBR), fluidized bed reactor (FBR), airlift reactor (ALR), upflow anaerobic sludge blanket (UASB) reactor, or any other suitable configuration. In UASB granular biofilm particles are used. This article demonstrates that reactor productivities in these reactors have been superior to any other reactor types. This article describes production of ethanol, butanol, lactic acid, acetic acid/vinegar, succinic acid, and fumaric acid in addition to wastewater treatment in the biofilm reactors. As the title suggests, biofilm reactors have high potential to be employed in biotechnology/bioconversion industry for viable economic reasons. In this article, various reactor types have been compared for the above bioconversion processes.
==== Body
Introduction
Biochemical reactors play an important role in the biochemical industry as the rate of reaction, ease, and length of reactor operation affect reactor productivities and hence process economics [1,2]. In order to employ a most appropriate reactor for an industrial operation, reaction rate should be high and the reactor configuration should be simple. Under optimized parameters such as pH, temperature, substrate, and medium components, reaction rate can be increased by increasing cell mass concentration in the reactor. There are two methods commonly used for increasing cell mass concentration inside the reactor; first, use of a permeable membrane to retain cells; and the other, use of immobilized cell technique. Membrane reactors allow passing of liquid, substrate, and product out of the reactor while retaining the cells. In these reactors, high cell concentrations can be achieved [3]. Unfortunately, for some processes such as waste water treatment, these reactors are not preferred due to their high cost and problems with fouling. Other processes where the relatively high cost of these reactors does not allow their use include production of large volume, low cost chemicals such as vinegar or acetic acid.
Other types of reactors that offer high reaction rates are immobilized cell reactors [4]. In these reactors, high cell concentrations are achieved by fixing them on various supports. Cells can be immobilized by three different techniques; namely, adsorption, entrapment, and covalent bond formation. Entrapment and covalent bond formation require use of chemicals that add to the cost of production and perhaps restrict further propagation or increase in cell concentration inside the reactor. The third technique is of natural origin as cells "adsorb/and adhere" to the support naturally and firmly [4-6]. This technique is called "adsorption" and has been used extensively in the literature to adsorb microbial cells. Table 1 shows a comparison of these techniques with the membrane reactors. It should be noted that some microbial cells leach out from immobilized cell reactors which require separation (leached out cells in reactor effluent) prior to product removal possibly by centrifugation.
Table 1 A comparison of different types of reactors with biofilm reactors
Reactor Type Comments
Membrane reactor
Advantages High productivities, high cell concentration can be achieved inside the reactor, clear permeates for further separation
Disadvantages Fouling with cells, cost prohibits their use in low cost large volume chemical production
Immobilized cell reactors
Covalent bond formation
Advantages High cell concentration may be achieved, high productivity
Disadvantages Cell growth inside matrix may be restricted, cells leach out of the matrix and hence centrifugation of effluent may be required, chemical may affect the cells
Entrapment
Advantages High cell concentration may be achieved, high productivity
Disadvantages Matrix often starts disintegration with time, cells leach out of matrix, centrifugation of reactor effluents is required for further separation
Biofilm
Advantages Comparatively high reactor productivities and high cell concentrations are achieved, reactors run longer and are economic to operate
Disadvantages Effluent centrifugation is required
In addition to being a natural process, adsorption can be performed in place, and economical adsorbents are available. Additionally, these reactors are simple in concept and construction and the immobilization process is economical. Adsorbed cells form cell layers on the support and cell mass grows inside the reactor over time [7]. These layers of cells are called "biofilms." Biofilms can be used in various types of reactors such as continuous stirred tank reactors (CSTRs), packed bed reactors (PBRs), fluidized bed reactors (FBRs), airlift reactors (ALRs), upflow anaerobic sludge blanket (UASB) reactors, and expanded granular sludge bed (EGSB) reactors etc. [4,7-11]. In these reactors, reaction rates are usually high as compared to the other types of reactors. On the laboratory, pilot plant, and industrial scale (some), these reactors have been very successful and examples include waste water treatment [12] and vinegar or acetic acid [13] production. In addition to these, other processes that have employed these biofilm reactors include ethanol, butanol, lactic acid, fumaric acid, and succinic acid production. Since they offer high reaction rates and are economical, this review becomes their subject matter. In the authors' view, this natural process of biofilm formation can be employed to economize production of various chemicals by fermentation on a large scale [13]. In biofilm reactors, cell concentrations as high as 74 gL-1 can be achieved [7]. In addition, the cell layers in bioparticles become highly active, thus contributing to the high reactor productivities. Within fluidized bed reactors, the biofilm particles are of various shapes (including spherical and irregular shapes) and these reactors can be operated for long periods of time. The amount of adsorbent that is used in these reactors is low, which also reduces the cost of the cell support. In biofilm reactors, reactor configurations can vary from a simple packed bed reactor to fluidized bed, UASB, and airlift reactors as described in this article.
Biofilm Formation
Various types of biofilms
In nature, biofilms exist primarily as complex multi-species communities of bacteria in which each species fills an ecological niche within the biofilm depending on its metabolism and morphology [14]. The nature of mixed culture biofilms is dependent on which species are present and what role each species fills. For instance, a single species may utilize anaerobic fermentation deep within one biofilm in one environment, but may utilize an aerobic metabolism in another environment in the presence of different neighboring biofilm species. Multi-species biofilms are important clinically as well as industrially. Clinically, biofilms are important as the source of persistent infections. They are responsible for dental caries and nosocomial infections, as well as a variety of other infections and diseases [15]. Industrially, biofilms are detrimental in many cases and beneficial in many others. For instance, natural biofilms can reduce heat transfer in heat exchangers and cooling towers [16], foul reverse osmosis membranes [17], and contaminate food processing equipment [18]. Multi-species biofilms are used industrially to achieve several aims including the treatment of wastewater for removal of organics [19,20] and heavy metals [21]. The presence of multiple species allows for the treatment of waste streams that are diverse in composition and that fluctuate in component concentration.
Single species biofilm are used to produce industrially important chemicals [22,23]. Such biofilms can exist in some situations and are important industrially, although in nature they are not the norm. For example, in nature, an immature biofilm that is resultant from the attachment and growth of a single cell may exist as a single species biofilm before incorporating other species. For chemical production, single species biofilms are important because they allow for control and maximization of desired products. In this case, a single species is inoculated into a sterile environment and allowed to form a biofilm before being used to produce a particular chemical product.
In industrial applications including wastewater treatment, usually two types of biofilms are employed, namely, biofilms that grow onto supports such as charcoal, resin, bonechar, concrete, clay brick, or sand particles, and biofilms that are formed as a result of flocs and aggregate formation. On the above supports, biomass grows all around the particles and the size of the biofilm particles grows with time usually to several mm in diameter. The density of the support particles is usually higher than the fermentation broth and for this reason bioparticles tend to remain in the lower section of the reactor. Another type of biofilm is where no support is used and cells form biomass granules and flocs that also grow in size with time. This type of biofilm is called granular biofilm and the reactor where this biofilm is used is called granular biofilm reactor. Granule formation may take from several weeks to several months. The cells produce extracellular polymeric substances (EPS) that binds the cells firmly in the form of flocs and aggregates. The most commonly used bioreactors that fall in this category are upflow anaerobic sludge blanket (UASB) reactors that are used to treat domestic and industrial wastewater anaerobically. Sponza [24] examined anaerobic granulation process in a UASB to remove tetrachloroethylene. In some cases expanded bed biofilm reactors have been used with granular biofilm particles that are called expanded granular sludge bed (EGSB) reactors.
Mechanism of biofilm formation
A biofilm is defined as a structured community of bacterial cells enclosed in a self-produced polymeric matrix and adherent to an inert or living surface [15]. In general, there are four stages to the development of a mature biofilm: initial attachment, irreversible attachment by the production of extracellular polymeric substances (EPS), early development, and maturation of biofilm architecture [14].
The life of a biofilm starts with the planktonic or free floating cell. In order for a planktonic cell to attach to a surface, it must first interact with the surface. Surfaces immersed in an aqueous solution usually acquire a surface charge which attracts and concentrates inorganic solutes, and charged or highly polar organic molecules. The concentration of cations, glycoproteins, proteins, and organic molecules at the surface can provide a relatively nutritious zone for bacteria compared to the bulk aqueous environment [25]. In addition, fluid flow in the boundary region near the surface can be considered negligible which allows bacteria to approach the surface. Once near the surface, it will either approach the surface by Brownian motion or move by chemotaxis towards the surface in response to the chemical concentration gradient [25]. When at the interface, the cell will form a temporary association with the surface or microbes already present on the surface [26].
After initial association with the surface, a planktonic bacterial cell can dissociate from the surface and resume the planktonic state or become irreversibly attached to the surface. Irreversible attachment involves the production of EPS. EPS serves to bind the cell to the surface and to protect it from the surrounding environment. EPS can be composed of polysaccharides, proteins, nucleic acids, or phospholipids. A common EPS produced by bacterial cells in biofilms is the exopolysaccharide alginate. In biofilm associated cells of Pseudomonas aeruginosa, transcription of algC, the gene involved in alginate production, was fourfold that in planktonic cells [27]. EPS provide protection to biofilm cells by providing a diffusive barrier to any toxic compounds that could harm the cells as well as a barrier to phagocytes and bacteriocides. The EPS can also represent a barrier to nutrients necessary for cell growth. Cells in the interior of a biofilm often show a much reduced rate of growth and cell division rate may be near zero [26,28]. The reduced growth rate is itself protective because uptake of toxic substances is also reduced. The presence of the EPS matrix may also serve as a spatial restrictor of cell growth and division.
Water and nutrient diffusion into the interior of a biofilm is highly limited. As biofilms mature, water channels can develop that allow water and nutrient access deeper into the biofilm. These channels partially relieve the diffusion limitation within the biofilm. The architecture of the biofilm develops in response to shear forces. In low shear environments, biofilms can form as thick mushroom-like masses. In high shear environments, biofilms may be flatter or form long strands [29].
A final stage that may occur in the life of a biofilm is reversion of part of the cells to the planktonic state. When cells living in biofilm take up nutrients, they channel much of that energy towards production of EPS rather than to cell growth and division. When nutrients become scarce, cells must escape the EPS matrix or be trapped in an unfavorable environment. Biofilm associated cells are able to produce enzymes capable of breaking down the EPS matrix in times of nutrient starvation. Pseudomonas fluorescens is able to produce an exopolysaccharide lyase under starvation conditions [30]. The enzyme serves not only to break down the polysaccharide matrix allowing cells to find nutrients elsewhere, but the degraded EPS can often be used as a food source for the nutrient deprived cells. In addition to cell detachment due to starvation or nutrient deficiency, there are other detachment processes such as abrasion, shear stress, sloughing and grazing. Providing detailed accounts of these processes is considered beyond the scope of this article.
Factors enhancing biofilm formation
Several parameters affect how quickly biofilms form and mature, including surface, cellular, and environmental factors. The surface onto which cells will attach has an important impact on biofilm formation. Rough surfaces tend to enhance biofilm formation [31]. Shear forces are lower near a rough surface, and there is a larger surface area to which cells can adhere. Porous materials also work well for biofilm formation. Shear forces are very low inside pores even under conditions where bulk fluid velocity is high. Pores provide a protected environment for cells to attach and grow. Porous materials such as brick and bonechar have been used to immobilize Clostridium cells used in biofilm reactors [22,32]. Biofilm formation also tends to increase with the hydrophobicity of the surface material [28]. Biofilms form much more rapidly on Teflon and other plastics than glass or metal. Possibly this is due to differences in hydrophobicity of the surfaces and ionic charges [28].
The amount of nutrients present in the medium can affect the rate of biofilm formation. Biofilms tend to form more readily in the presence of ample nutrients [33]. One function of the biofilm is to anchor cells in a friendly, nutrient rich environment. Phosphorus is a particularly important nutrient. Cells saturated with phosphate have a higher tendency to flocculate and adhere due to their increased hydrophobicity, while those cells depleted in phosphate are more hydrophilic and less likely to adhere [34].
Temperature can have an effect on biofilm formation. Temperatures at the high end of a culture's growth range can enhance biofilm formation. Depending upon the species involved, high temperature increases the rate of cell growth, EPS production, and surface adhesion, all of which enhance biofilm formation [25].
Cellular factors may affect biofilm formation. A hydrophobic cell will be more able to overcome the initial electrostatic repulsion with the solid surface and adhere more readily. The presence of fimbriae, proteinaceous bacterial appendages high in hydrophobic amino acids, can increase cell surface hydrophobicity [28,35]. Flagellated cells show increased ability to attach to surfaces. Flagellar motility may serve to overcome initial electrostatic surface repulsion.
Calculations and data presentation
In a continuous process, productivity (gL-1h-1) is calculated as the product concentration in gL-1 liquid multiplied by the dilution rate (h-1). In a batch process, productivity is calculated as the product concentration in gL-1 liquid divided by the fermentation time (h). Specific productivity (h-1) is calculated as productivity (gL-1h-1) divided by cell or protein concentration (gL-1). Dilution rate (feed flow per reactor volume per h) can be based on total volume of the continuous reactor or void volume. In fully or partially packed bed reactors, void volume is total reactor volume minus the volume occupied by the cell support. For a particular flow rate, dilution rate based on void volume is higher than based on the total reactor volume. In this article, reactor productivities based on both total reactor volume and void volume have been reported as mentioned by the different authors. The reader is advised that it is difficult to correlate/compare the two productivities (based on total reactor volume or void volume) unless void volume fraction (void volume/total volume) is given along with the flow/feed rate. Residence time (h) in the reactor can be calculated by inversing the dilution rate (h-1).
Types of biofilm reactors
Biofilm reactors can be assembled in a number of configurations including batch, continuous stirred tank (CSTR; including agitating continuous reactors, and rotary continuous reactors), packed bed (PBR), trickling bed (TBR), fluidized bed (FBR), airlift reactors (ALR), upflow anaerobic sludge blanket (UASB), and expanded bed reactors. The operation of these reactors changes from reactor to reactor. In a batch biofilm reactor, the immobilized cells have to be utilized for repeated batches. However, it is likely that during the late stationary phase of chemical production the culture would experience inhibition thus reducing productivity. Also, in a batch reactor, productivity would be reduced due to downtime necessary to fill and empty the reactor. If the reactor is packed with biofilm particles, some cells may die or become inactive due to lack of feed during emptying and filling of the reactor. As a result, it is viewed that batch reactors are not practical for biofilms.
In a CSTR feed medium is fed to the reactor and product is withdrawn at the same rate as feed. They are stirred using a mechanical device such as impeller. CSTRs cannot be packed with the adsorbent support covered by biofilms as no agitation can be provided in that case. However, they can be used if fibrous bed support is used for adsorption of cells. In that case, cells can grow and form a biofilm on the fibrous bed. In such a case agitation can be provided. This type of system was used for the production of butanol [36] and lactic acid [37] in continuous operation with a constant feed and a constant effluent from the reactor. In some cases, there may be excessive growth on the surface of the fibrous bed, and the cell layers may be sheared off the support. This type of fibrous bed biofilm CSTRs are called as agitating continuous reactors. Another type of CSTRs called rotating CSTRs have same length/diameter (L/D) ratio as in CSTRs. The rotating CSTRs are placed horizontally (lengthwise) and are rotated along the horizontal axis.
PBRs are different types of reactors as they are packed with suitable support material followed by inoculation with the culture to form biofilm. The reactor is supplied with a feed that is not deficient in nutrients. Depending on the culture, nutrients, and support, biofilm formation may take a few to several days. Such reactors are usually fed at the bottom, thus getting product at the top of the reactor. However, these reactors are prone to blockade due to excessive cell growth. In C. acetobutylicum/C. beijerinckii biofilm packed bed reactors, reaction rates up to 45 times that of the batch (control) reactors have been obtained [38,39].
TBRs are different from PBRs as they (TBRs) are fed at the top of the reactor thus obtaining product at the bottom. However, in such reactors some of the biofilms may not get sufficient feed thus affecting reactor efficiency/productivity adversely. Also, in gaseous fermentations gas may occupy significant space in the reactor and may form stagnant pockets. This also may affect the efficiency of the reactor. In anaerobic waste water treatment and acetic acid production, these reactors have been used at large scale successfully.
FBRs have played a successful role in the degradation of toxic phenolic chemicals [40-43] and butanol production [39,44]. In these reactors, cell growth occurs around the adsorbent particles. Formation of active biofilms around the particles and accumulation of sufficient biomass in the reactor may take from 2 to 4 weeks. A major advantage in these reactors is that they can be operated for much longer periods than PBR or CSTRs (with fibrous bed). These reactors do not block due to excessive growth. In these reactors butanol production was increased by approximately 40–50 times that of the batch reactors. These reactors have been operated successfully for longer than 4 months in continuous operation (Unpublished data, Qureshi and Maddox).
Airlift reactors contain two concentric tubes, a riser (an inner tube) and a downcomer (an outer tube). In these reactors, mixing is achieved by circulating essentially air at the bottom of the reactor. As a result of force applied by the air (at the bottom of the inner tube), the liquid in the inner tube moves up which then overflows (the inner tube) downward thus creating eddies to mix the liquid. In some of the airlift reactors downcomer is replaced with an external loop to circulate fermentation broth. Such reactors where air is replaced by an anaerobic gas are called gaslift reactors.
Upflow anaerobic sludge blanket (UASB) reactors (contain granular biofilm particles) are used for anaerobic treatment of wastewater/industrial effluents. As the name suggests, the flow in these reactors is in upward direction. At the top of the reactor provisions are made for gas/es to escape and sludge particles to settle to the bottom part of the reactor. Reactor effluent is removed from the top of the reactor. UASB reactor was developed by Lettinga et al. [9]. Fig. 1 shows a schematic diagram of various reactors and biofilm particles.
Figure 1 Schematic diagrams of various types of biofilm reactors and biofilm particles.
Biofilm reactors in biological wastewater treatment
The application of biofilm technology in wastewater treatment originated from the industrial operation of trickling filters in the early 1880s in Wales, Great Britain [45]. Biofilm processes in wastewater treatment can be divided into two categories: namely (1) the fixed-medium systems where the biofilm media are static in the reactors and the biological reactions take place in the biofilm developed on the static media, and (2) the moving-medium systems where the biofilm media are kept continually moving by means of mechanical, hydraulic, or air forces [46]. The moving-medium systems include rotating biological contactors, moving-bed biofilm reactors, vertically moving biofilm reactors, and fluidized bed biofilm reactors; while the fixed-medium systems include trickling filters and biological aerated filters [46]. Rotating biological contactors (RBC) have been widely used in biological treatment of wastewater for reducing chemical oxygen demand (COD)/biological oxygen demand (BOD) [47,48] and nitrification/denitrification purposes [48,49]. Rotating biological contactors treat wastewater streams using a thin biofilm of aerobic microorganisms on rotating cylinders or biodiscs. The rate of rotation is selected to provide optimum contact of the waste stream with the biofilm for efficient oxygen transfer and bioactivity.
It should be noted that before a biofilm-based treatment system is to be considered for the treatment of wastewater, it is necessary to determine whether the naturally occurring microorganisms are able to produce biofilms, while simultaneously reducing the COD of the wastewater, or if there is a need to inoculate the reactor with external bacterial strains. The most commonly used rotating biofilm contactor is the rotating biodisc and its various modifications [48]. For the treatment of high strength wastewater, gentle aeration of the liquid phase has been shown to improve the COD reduction of the system by about 40% [50]. In a related study, Kargi and Eker [48] have shown that a rotating-perforated-tube biofilm reactor is effective in COD removal from synthetic wastewater composed of diluted molasses, urea, KH2PO4, and MgSO4. The liquid phase in the tank was not aerated, (the total biofilm surface area (A) was 1.34 m2), and the rotation speed of the tubes was 5 rpm [48].
In some instances, thermophilic aerobic systems have been employed to biodegrade the wastewaters of high strength, and tremendous COD reductions have been reported in both laboratory and pilot scale experiments [51,52]. However, the thermophilic systems exhibited poor bacterial flocculation characteristics due to the dispersed growing microorganisms (no biofilm formation) which made bacterial separation from the treated effluent difficult [53,54].
The fluidized bed biofilm reactors (FBBR; also called as FBR) (in which particles move up and down within the expanded bed in the well defined zone of the reactor) have been used for more than two decades for treating industrial wastewater [55,56]. Immobilized bacterial systems configured as fluidized bed biofilm reactors (FBBRs) offer some technical advantages. Since chemical wastes are injected into the recycle, toxic chemicals are immediately diluted, which make the microorganisms more resistant to direct chemical toxicity than many conventional treatment systems. In addition, since FBBRs are usually oxygenated by supplying air into the recycle loop, a high level of microbial activity may be supported with minimal air stripping of volatile chemicals.
Jesis and Owen [57] studied FBBR, and they found that the use of small, fluidized media enabled the FBBR to retain high biomass concentrations and, thereby, operate at significantly reduced hydraulic retention times. In pilot scale operations carried out by Jesis and Owen [57], they reported that when the volatile solid concentrations were between 30,000 to 40,000 mgL-1 during denitrification operation, 99% of influent nitrates could be removed under hydraulic retention times as low as 6 min. In a related study, Rabah and Dahab [56] during their evaluation of the use of fluidized-bed biofilm reactors for nitrate removal concluded that the FBBR system is capable of handling an exceptionally high nitrate nitrogen concentration of 1000 mg(N)L-1 with very high removal efficiency, up to 99.8%. The authors noted that higher denitrification rates can be achieved at relatively low superficial velocities because it is possible to maintain high biomass concentration at lower velocities. However, there is a minimum practical velocity below which agglomeration of media would occur and the process may fail [56]. The efficiency of the FBBR can be up to 10 times greater than that of the activated sludge system and typically occupies 10% of the space required by stirred tank reactors of similar capacities [56]. Higher biomass concentration in the FBBRs (40, 000 mgL-1) compared to 3000 mgL-1 in the activated sludge have been shown to be the reason for the greater efficiency [58].
Anaerobic treatment of wastewater in fluidized bed reactors is another area that has been studied extensively [59]. In this article, Iza [59] presented theoretical basis for design and operation of a fluidized bed reactor for anaerobic treatment of wastewater. The anaerobic fluidized bed technology offers a number of advantages for treating wastewater including high concentration of biomass attached to the dense support that makes it possible to operate them at high dilution rate without cell washout. In these reactors no plugging, gas hold-up or channeling occurs.
Prior to the development of UASB, interest in anaerobic treatment systems in wastewater treatment was scarce [11]. Interestingly, the development of UASB saw a significant increase in anaerobic removal of various chemicals from the wastewater using these reactors. The examples include anaerobic removal of pentachlorophenol [60], nitrogen removal [61], dechlorination using Dehalospirillum multivorns [62], anaerobic treatment of municipal solid leachate [63], and starch degradation [64]. More studies on this subject have been reviewed in Veeresh et al., [42] (anaerobic treatment of phenol and cresols in UASB reactors). The success of the UASB concept relies on the establishment of a dense sludge bed in the bottom of the reactor which is usually a result of microbial growth and incoming sludge.
Seghezzo et al., [11] reported that in a pilot plant UASB reactor, internal mixing was not optimal for treating sewage (4–20°C) which produced dead space and hence reduced process efficiency. In order to improve the process efficiency, an adequate influent distribution was sought. The use of effluent recirculation in combination with a taller reactor (a larger height to diameter ratio) resulted in the expanded granular sludge bed (EGSB). Usually expended bed reactors, as opposed to EGSB, have biofilm that is adsorbed onto support particles. An example of expanded bed reactor is that of Tsuno et al., [43] who degraded pentachlorophenol (PCP) in a biological expanded-bed reactor anaerobically. In this reactor the granular activated carbon was used as a support.
Generally, total nitrogen removal from domestic or industrial wastewater streams is achieved in two steps: microbial nitrification of ammonium (aerobic process) followed by denitrification (anaerobic process) or reduction of formed nitrate to nitrogen. This conventional method employs a sequence of aerobic and anoxic processes in order to provide the two different environmental conditions [65]. However, studies have shown that these two important steps can occur simultaneously in one reactor in a process called simultaneous nitrification and denitrification (SND). In SND process, nitrification is restricted to the outer oxic zone of formed microbial flocs, whereas denitrification occurs predominantly in the inner anoxic zones [65]. To test the hypothesis that SND is a physical phenomenon, Pochana and Keller [66] carried out experiments to determine the effect of floc size on SND. Typical floc sizes as measured in their experiments were 50 – 110 μm, which is large. Such large floc sizes could create an anoxic zone inside the flocs leading to denitrification. Pochana and Keller [66] concluded that a substantial anoxic mass fraction exists in the center of the biomass floc resulting from an oxygen diffusion limitation into the floc. The rates of nitrogen removal (coupled nitrification-denitrification process) or productivities/specific productivities are shown in Table 2. The sequencing batch and single activated sludge flocs reactors require some O2 to effect nitrification (NO3-/NO2- generation), which is the precursor for denitrification process.
Table 2 Productivities of different reactors employed for nitrification and denitrification/during the pretreatment of domestic or industrial wastewater streams.
Reactor Type Removal rates (Productivities or specific productivities) Reference
Nitrification Denitrification
Activated sludge flocs (Single reactor) 24 μmol N g MLSS-1 h-1 6 μmol N g MLSS-1 h-1 [65]
Sequencing batch 19 mg NH3 L-1h-1 13.5 mg N L-1h-1 [66]
Chemostat (Continuous) 5.6 μmol NH3 h-1 mg protein-1 5.6 μmol N h-1 mg protein-1 [67]
Chemostat (Continuous) 2.6 μmol NH3 h-1 mg protein-1 4.1 μmol N h-1 mg protein-1 [68]
Continuous 250 μmol NH3 L-1h-1 400 μmol N L-1h-1 [69]
MLSS = Mixed liquor suspended solid
In contrast to previous view that denitrification occurs under anaerobic conditions [67] it (denitrification) has been shown to occur under aerobic conditions with a wide range of bacteria [68,70]. Robertson and Kuenen [71] observed that under fully aerobic conditions, Thiosphaera pantotropha carries out the following reactions sequentially and simultaneously, in the presence of a suitable electron donor such as acetate.
This implies that the organism can convert ammonia into nitrogen gas without intermediary accumulation of nitrite. Investigating the reason why this organism denitrifies under aerobic conditions, Robertson and Kuenen [72] demonstrated that denitrifying enzymes were present even when the organism was growing aerobically without nitrate.
The last decade has witnessed an increased interest in membrane bioreactors for wastewater treatment [73,74]. The membrane-aerated biofilm reactor (MABR), whereby the biomass is immobilized on membranes through which oxygen is supplied seems to be the most promising design. Results from studies with MABRs have been reported for the degradation of phenol [40], chlorophenols [41], xylene [75], and ammonia [76]. Stripping losses of volatile organic compounds are minimized, and the oxygen partial pressure in the gas compartment allows easy control of oxygen penetration into the biofilm; the dissolved oxygen gradient across the membrane and the biofilm offers an ideal environment for aerobic strains, and foaming due to surfactants can be prevented [74]. For high strength wastewaters, the possibility of enhanced oxygen penetration depths makes MABRs an attractive option for pollutant biodegradation. However, Casey et al. [74] reported that an excessive growth of biofilm is frequently observed. Therefore, biofilm growth should be controlled when operating this reactor. Although a significant amount of work has been performed on the use of biofilm reactors in wastewater treatment, it is beyond the scope of this article to discuss this work in greater detail.
Biofilms for gas and odor treatment
Traditionally, industrial waste gases have been treated by physico-chemical methods known as adsorption, scrubbing, condensation, and oxidation processes [77]. Biological waste gas treatment is an attractive and environmentally friendly alternative to physico-chemical methods. Industrial waste gases can serve as energy or carbon sources for microbial metabolism. In addition, inorganic waste gases (H2S, NH3) may be treated directly by employing autotrophic microorganisms which have the ability to utilize CO2 as a carbon source for anabolism [78]. Koe and Yang [79] during their evaluation on how to drastically reduce or eliminate the impact of air polluting emissions from wastewater treatment plant suggested that open sources of odorous emissions such as inlet works, primary sedimentation units, aeration tanks, final clarifiers, sludge processing units, and wastewater channels should be covered up and the odorous air be treated before discharging to the ambient atmosphere.
The biofilter, trickling biofilter, and bioscrubber are three major bioreactor designs frequently employed for the treatment of waste gas [78]. A biofilter consists of a filter-bed composed of a carrier (sawdust, compost, dry wastewater sludge, etc.) for the active microorganisms and as nutrient source [77]. Biofilters operate by facilitating the transfer of odorous gas from waste air blown through the biofilters into biofilms around particles of biofilter medium in which bacteria, fungi and other microorganisms are immobilized. On the other hand, waste gas treatment in trickling biofilters involves use of a biological filter continuously fed with a liquid medium and packed with a synthetic carrier on which biofilms grow [77]. Trickling biofiltration has been used, especially outside the United States, for removal of odorous waste gases such as H2S [80]. Several species of microorganisms can oxidize hydrogen sulfide to form odorless sulfuric acid. Thiobacillus thiooxidans is capable of oxidizing H2S at low pH [81]. For effective H2S odor control, an ideal habitat for the growth of sulfide-oxidizing bacteria should be created and competing microbes which normally predominate in aerobic treatment processes should be excluded. De Beer et al., [82] demonstrated that the channels surrounding the cell clusters could increase the supply of oxygen and other nutrients to cells within the biofilm, thus relating structure to function [14]. The biofilm structure appears to be largely determined by the production of slime-like matrix of extracellular polymeric substances, which provide the structural support to the biofilm [14]. The structure of biofilms is largely determined by a number of biological factors such as microorganism growth rate, motility, cell signaling, and the production of extracellular polymeric substances. The physical growth environment may also play a significant role in the determination of the biofilm structure [14], and hence the efficiency. However, excessive biofilm development can lead to clogging of the filter-bed of the reactor [78].
Biomass growth and biofilm development can be limited by reducing nutrient supply although this may decrease reactor performance since higher biomass growth shows higher substrate consumption rates [83]. Therefore, it is important to find a balance between excessive biomass growth to prevent biofilter clogging and the odorous gases removal efficiency. Furthermore, waste gases that are characterized by high concentrations of water-soluble pollutants can be treated with bioscrubber. The bioscrubber consists of two reactors. The first reactor is an absorption column where pollutants are absorbed in a liquid phase. The liquid phase goes to the second reactor, which consists of a filter with an activated carbon medium that supports microbial growth. The high bioactivity in the bioscrubber enhances conversion of waste gases into nonhazardous and less odorous compounds. The effluent leaving the bioscrubber can be re-circulated to the absorption column; this technology allows for good gas cleaning when the gaseous pollutants are highly water soluble [77].
Ottengraf [78] reported that the rate of mass transfer of a given compound to be removed or deodorized is determined by the product of the overall mass transfer coefficient, the total contact area in the column, and the average driving force. Therefore, the absorption of a compound will be higher if its concentration in the wastewater is low and its solubility in water is high [78]. The control of operating parameters to the microorganisms in these bioreactors can sometimes be challenging.
Production of industrial chemicals in biofilm reactors
Biofilms and biofilm reactors in ethanol production
Bland et al. [84] produced ethanol in an attached film expanded bed bioreactor of Zymomonas mobilis. The cells of Z. mobilis were adsorbed onto vermiculite and the culture formed an active biofilm. Based on the total volume of the reactor, a productivity of 105 gL-1h-1 was obtained at a dilution rate of 3.6 h-1. Usually, in a control batch or free cell reactor a productivity of <4 gL-1h-1 is achieved. The increased/enhanced productivity reported here is due to the formation of active biofilm onto the adsorbent.
Adsorbed cells of Saccharomyces cerevisiae were used in a packed bed continuous bioreactor to produce ethanol from molasses [4]. The cells were immobilized onto a support of natural origin, possibly sugarcane bagasse. It has been reported that the cells were immobilized by natural mode, which is likely to be adsorption. The amount of cells that was adsorbed onto this support was 0.13 gg-1 support. In this biofilm reactor, the authors reported a productivity of 28.6 gL-1h-1 as compared to 3.35 gL-1h-1 in a free cell continuous process. The dilution rates in the biofilm reactor and free cell continuous system were 0.47 h-1 and 0.65 h-1, respectively. Although immobilized cell reactors (such as this biofilm reactor) are typically operated at higher dilution rates than the free cell continuous reactors, it is not clear why the authors used a lower dilution rate in the biofilm reactor. Since carbon utilization for newly growing cells was reduced, product yield was improved as compared to a batch reactor.
Since ion exchange resins have charge on them, bacterial cells can be adsorbed onto the resins thus forming biofilm layers. This concept was employed by Krug and Daugulis [85] to produce ethanol in high productivity reactors using Z. mobilis. To find a suitable adsorbent, 10 ion exchange resins, activated carbon, and ceramic chips were examined. A cationic macroreticular resin was shown to be the most efficient adsorbent to immobilize cells of Zymomonas mobilis. The immobilized cells were used in a continuous column and 100 gL-1 glucose was fed to the reactor. As a result of formation of biofilm, the reactor productivity was measured at 135.8 gL-1h-1 (void volume based productivity, Pdv = 377.4 gL-1h-1). The reactor stopped working due to excessive cell growth and plugging after a period of 200 h of operation.
Other reports on ethanol production in biofilm reactors are those of Kunduru and Pometto [86] and Demirici et al. [8]. Kunduru and Pometto [86] studied ethanol production in continuous reactors using biofilm supports of polypropylene or plastic composite. Employing a culture of Z. mobilis and a bacterial support of polypropylene, a staggeringly high productivity of 536 gL-1h-1 was obtained at a dilution rate of 15.36 h-1. In a control free cell fermentation, a productivity of 5 gL-1h-1 was obtained at a dilution rate of 0.5 h-1. The biofilm reactor was fed from the top, thus collecting product at the bottom of the reactor.
Kunduru and Pometto [86] used another biofilm reactor of S. cerevisiae adsorbed onto a plastic composite support and reported a productivity of 76 gL-1h-1 at a dilution rate of 2.88 h-1. The reactor productivity in a control reactor was 5 gL-1h-1 at a dilution rate of 0.5 h-1. Unlike the above biofilm reactor, S. cerevisiae biofilm reactor was fed at the bottom, and the product was obtained from the top. It is suggested that for a proper comparison both the reactors should have been fed in the same direction.
In order to enhance biofilm formation, Demirici et al. [8] developed a new support material for the growth of S. cerevisiae. A mixture of ground soybean hulls (or oat hulls), complex nutrients, and polypropylene was extruded at high temperature into disks and rings. It is likely that heat sensitive nutrients were inactivated during extrusion. Also, polypropylene film may have covered the nutrients, thus making them unavailable to the culture for cell growth. Since no data have been provided on the time period of formation of biofilm or thickness of biofilm, it is difficult to compare this support with other supports.
In a more recent study, Qureshi et al. [87] produced ethanol in a biofilm reactor of genetically engineered Escherichia coli from xylose. The biofilm was formed on clay brick particles, and the reactor was operated continuously for 103 days. The reactor was operated at various flow rates, and reactor productivity was found to be improved compared to a free cell batch process. Table 3 compares ethanol productivities obtained in biofilm reactors of various cultures.
Table 3 A comparison of production of ethanol in adsorbed cell biofilm reactors
System/Support Reactor Type Culture Productivity [gL-1h-1] Reference
Biofilm Reactors
Resin Packed bed Z. mobilis 135.8 (PdT), 377.4 (Pdv) [85]
Vermiculite Packed bedd Z. mobilis 105.0 (PdT), 210 (Pdv) [84]
Sugarcane bagassea Packed bed S. cerevisiae 28.6b [4]
Polypropylene Packed bede Z. mobilis 536c [86]
Plastic composite Packed bed S. cerevisiae 76c [86]
Cell Recycle CSTR Z. mobilis 200b [88]
Batch/Continuous suspended cell (Control)
Continuous CSTR Z. mobilis 5.0b [86]
S. cerevisiae 5.0b [86]
Continuous CSTR S. cerevisiae 3.35b [4]
a: The support was reported as an adsorbent of natural origin (perhaps sugarcane bagasse)
b: Not reported whether based on total reactor volume or void volume
c: Not reported (possibly based on void volume)
d: Cone shaped
e: Trickling packed bed
PdT – Productivity based on total reactor volume
Pdv – Productivity based on reactor void volume
Biofilms and biofilm reactors for butanol production
Butanol is an important industrial chemical that can be produced from a number of carbohydrates using a number of microbial cultures. Butanol can be used as a fuel and has higher/greater energy content than ethanol. Production of butanol has been investigated in batch, fed-batch, free cell continuous, immobilized cell continuous, and cell recycle continuous reactors [1]. Continuous immobilized cell and cell recycle reactors offer higher productivities than batch and free cell continuous reactors. In addition to achieving a high productivity, a major advantage of immobilized cell technology is that there is no cell washout at high dilution rates.
Adsorption is a technique which does not require any chemicals for cell immobilization and can be easily performed inside the reactor. In order to immobilize cells of Clostridium acetobutylicum, the reactor is packed with an adsorption support followed by inoculation with the culture. The adsorption process varies from 2–3 days to weeks depending upon the culture, support, and the reactor. The culture forms cell layers (biofilm) on the support [5,6,22].
An early report of adsorption of cells of C. acetobutylicum for the production of butanol was that of Forberg and Haggstrom [5]. These authors used beechwood shavings to adsorb cells. The reactor was fed continuously with a glucose solution (and nutrient dosing). Over a period of time, an active biofilm was formed on the wood shavings, and a reactor productivity as high as 1.53 gL-1h-1 was observed (compared to <0.1–0.35 gL-1h-1 in control batch fermentation). This work was followed by experiments examining the production of butanol in adsorbed cell biofilm reactor of C. acetobutylicum from whey permeate [6]. It should be noted that biofilm formation on this support was quick, and a reactor productivity of 4.5 gL-1h-1 was observed, which was superior to any previously reported butanol production system. Following these reports, Welsh et al. [89] investigated the use of a number of adsorption supports for butanol production by C. acetobutylicum in batch and continuous systems. The adsorbents used were coke, kaolinite, and Gel White (a montmorillonite clay). Coke was reported to be superior to other supports for adsorption. A maximum concentration of acetone butanol ethanol (ABE) in the effluent of the reactor was reported to be 12 gL-1 at a dilution rate of 0.1 h-1, thus resulting in a productivity of 1.2 gL-1h-1.
Following above reports, an intensive study was performed on the adsorption of C. acetobutylicum on a number of supports and biofilm formation (Table 4, 5). It has been observed that C. acetobutylicum and C. beijerinckii form visual biofilm layers in 2–4 days (in packed bed reactors), and reactors become productive after 4th day of continuous operation. The techniques of adsorption and reactor operation have been reported previously [6,38,89]. It has been observed that not all the supports are suitable for adsorption (Table 4). It has also been observed that during biofilm formation onto bonechar, the culture produces higher concentration of polysaccharide between day 2 and 4. During this period, up to 2.04 gL-1 polysaccharide production was observed as opposed to 0.95 gL-1 during day 5–30 (Fig. 2). As described in the previous section, the cultures that were used for adsorption for butanol production have flagella, which perhaps help bring the cells closer to the surface of support. In addition, charge on the support and cell is likely to aid in initial adsorption or bringing the cells closer to the support surface.
Table 4 Biofilm formation characteristics of Clostridium acetobutylicum/C. beijerinckii onto various supports
Support Characteristics
Bonechar C. acetobutylicum culture
- Adsorption is quick
- Biomass layers (biofilms) become visible in 3–4 days time
- Between day 2 and 4, the culture produces polysaccharide in high concentrations (2.04 gL-1 broth as compared to 0.95 gL-1 broth from day 5 to 30)
- Once initial layers appear, biomass accumulation is quick
- Desorption does not occur at high dilution rates
- < than 25% cells were desorbed when adsorbed cell particles were agitated at 200–300 rpm (in shake flasks on shaker) at pH 2.7 for 18–24 h at 30°C
- During initial stages (2–4 days) the culture produced high concentrations of acids (~6–9 gL-1) followed by becoming solventogenic
- During solventogenic stages fluctuations in solvent concentrations were less
Glass beads C. acetobutylicum culture
- Biomass accumulation takes much longer than bone char
- During initial stages (2–4 days) higher amount of polysaccharide production does not occur
- Cells do not stick to the support as firmly as onto bonechar
- Reactor produces <20% solvents as compared to bonechar adsorbed cells
- Reactors are not stable as solvent concentration fluctuates
- Cells can easily be washed off
Glass wool, Polypropylene tow, and stainless steel wire balls C. acetobutylicum culture
- <20% biomass accumulated than in bonechar packed reactor
- Cells do not stick to the support firmly and can be desorbed easily
- Reactors are not stable and poor solventogenesis occurred
Clay brick (Ref. 38) C. beijerinckii culture
- Cells stick firmly as in case of bonechar and reactors were solventogenic
Table 5 Production of solvents in packed bed biofilm reactors of C. acetobutylicum/C. beijerinckii
Culture/support Maximum Solvent [gL-1] Maximum productivity
[gL-1h-1] Accumulated biomass [gL-1 reactor vol] Biomass accumulation [gg-1 support]
C. acetobutylicum
Bonechar 9.3 (0.30) 6.50 (1.5) 74.0 0.087
Glass beads 3.0 (0.31) 0.93 (0.31) 65.0 0.044
Glass wool 3.0 (0.10) 0.30 (0.10) 3.1 0.050
Polypropylene tow 2.3 (0.25) 0.58 (0.25) 0.8 -
Stainless steel wire balls 2.0 (0.07) 0.15 (0.07) 1.0 -
C. beijerinckii
Clay brick [Ref 38] 7.9 (2.00) 15.8 (2.00) 73.7 0.093
Numbers in bracket are dilution rates (h-1) at which solvent and productivity were obtained
- Values not calculated
Figure 2 Production of polysaccharide and accumulation of cell mass during the initial 3 days of adsorption of cells of C. acetobutylicum onto bonechar for the production of butanol from whey permeate in a packed bed reactor. 1. Initiation of excessive production of polysaccharide; 2. Maximum growth and attachment starts; 3. Biofilms become visible and polysaccharide production continues.
Some supports accumulated more cell concentration (C. acetobutylicum) and were more solventogenic than the others (Table 5). At this stage, we are not aware what makes some supports better than others for biofilm formation and cell accumulation. From some supports it was easier to wash away the cells while from others such as bonechar and clay brick it was more difficult (Table 4). During our studies on butanol production, it was observed that approximately 0.9–1.0 gL-1 cells were present in the effluent [7,90] of the reactor. We have demonstrated that the cells that are present in the effluent of the reactor are those that grew on the surface of the support, rather than those that grew in liquid medium inside the reactor [90], suggesting that a tremendous amount of activity occurs on the surface of the biofilm in C. beijerinckii/C. acetobutylicum cultures. The thickness of biofilm that is formed in C. acetobutylicum or C. beijerinckii cultures can range from few cell layers to as many as 35 or more. Figure 3 shows adsorbed cells and biofilm formed by C. acetobutylicum onto bonechar. Similar observations on biofilm formation were observed for C. beijerinckii [38].
Figure 3 Scanning electron micrograph of adsorbed cells of C. acetobutylicum P262 onto bonechar. a) bonechar (magnification 5500); b) adsorbed cells onto bonechar (magnification 2200); c) transmission electron micrograph of adsorbed cells (magnification 2300). Similar figures (3b, c) with different magnification were published previously in the following article: Qureshi N, Paterson AHJ, Maddox IS: Model for continuous production of solvents from whey permeate in a packed bed reactor using cells of Clostridium acetobutylicum immobilized by adsorption onto bonechar. Appl Microbiol Biotechnol 1988, 29:323–328. Figure 3 is reprinted with permission from Springer, Germany (see above article).
Intensive research has been done on butanol production in various types of reactor systems [1,39,91,92]. The biofilm reactor systems that have been used for butanol production include vertical packed bed reactor (PBR), horizontal PBR, compartmentalized reactor, double series reactors, and FBR. The PBRs and FBRs are different in the sense that FBR is started with support <10% of its volume while packed beds are filled up to 90% of their volume. In packed bed reactors, as cell growth occurs, they are often blocked due to excessive cell growth while in FBRs this does not occur. In FBRs, the bed is fluidized either by recycling fermentation broth, using anaerobic gases (N2 or CO2 & H2 in case butanol fermentation) or air (for other aerobic systems). Cell growth occurs all around the support particles and over a period of time the volume of biofilm particles becomes many fold greater than the support particle (Fig. 3, 4). It should be noted that in FBRs cell growth occurs on the particles in spite of broth's high flow rates [44]. In this fluidized bed reactor liquid flow velocity of the order 40–60 ms-1 was maintained. The reader is advised that despite such a high flow velocity, the culture maintains its growth as a biofilm. We have not calculated the shear rate on the biofilm particles. The reactor was used for the production of butanol from whey permeate in continuous operation for >4 months (unpublished results – Qureshi & Maddox). Newly adsorbed C. acetobutylicum cells onto bonechar grow in an exponential manner and accumulation of biomass continues with time. Figure 5 shows a picture of a fluidized bed reactor used for the production of butanol from whey permeate.
Figure 4 Photographs of biofilm particles of C. acetobutylicum P262 used in a fluidized bed reactor for the production of butanol from whey permeate. A) bonechar particles; B) biofilm particles after growth (bonechar particles are covered with biofilm layers).
Figure 5 A photograph of a fluidized bed bioreactor (inside volume 450 cm3) used to produce butanol from whey permeate using C. acetobutylicum P262.
Among the various types of reactors used for butanol production, adsorbed cell biofilm reactors (cells adsorbed onto bonechar and clay brick) offered the highest reactor productivities. The reactor productivities that have been achieved in these reactors ranged from 6.5 [39] to 15.8 [38] gL-1h-1 (as compared to 0.10–0.38 gL-1h-1 in batch reactors). Membrane cell reactors also offer high productivities (6.5 gL-1h-1) [93,94]; however, biofilm reactors were superior to these reactors (Table 5). Of the various supports tested, bonechar and clay brick were found to be most suitable, and strong biofilms were formed on these supports. C. acetobutylicum was adsorbed onto bonechar while C. beijerinckii was adsorbed onto clay brick. Attempts were made to desorb the adsorbed cells of C. acetobutylicum. In order to achieve this, 100 g bonechar with adsorbed cells (25 days old reactor) was transferred to a 500 mL conical flask. The pH of the solution/reaction mixture was adjusted to 2.7, and the mixture was placed on a rotary shaker at 250 rpm for 18 to 24 h. After this period <30% cells were desorbed from the bonechar.
Biofilms in 2,3-butanediol production
In an attempt to improve reactor productivity in 2,3-butanediol fermentation, Maddox et al. [23] immobilized cells of Klebsiella pneumoniae on to bonechar. The cells of K. pneumoniae were adsorbed in a similar manner as C. acetobutylicum [6]. During the 2,3-butanediol fermentation, a productivity of 11.7 gL-1h-1 was obtained, which was the highest reported productivity. Prior to this work Shazer and Speckman [95] reported a productivity of 1.04 gL-1h-1 in 2,3-butanediol fermentation using Bacillus polymyxa in a membrane cell reactor. This work clearly demonstrated that bonechar adsorbed cells of K. pneumoniae result in superior productivities. Table 6 compares reactor productivity achieved in biofilm reactor as compared to various other reactor types. It should be noted that although high reactor productivity was obtained in the adsorbed cell reactor, cells did not adsorb on to bonechar as strongly as C. acetobutylicum. Rather, cells were entrapped in between bonechar particles. However, it is anticipated that there were a significant amount of cells sitting on the surface of bonechar as bonechar surface area was large. At the end of fermentation, it was observed that unlike cells of C. acetobutylicum, K. pneumoniae cells were washed away easily. It is not known whether K. pneumoniae cells produce polysaccharide which adds/facilitates adsorption of cells to the surface of bonechar or other surfaces. Even though K. pneumoniae cells do not form firm layers of cells, these reactors are still highly productive.
Table 6 A comparison of 2,3-butanediol productivity in a packed bed biofilm reactor with productivities in other reactor types
Reactor Type Culture Substrate Productivity [gL-1h-1] Reference
Biofilm, continuous K. pneumoniae Whey permeate 11.70 [23]
Batch (control) A. aerogenes Glucose 1.10 [96]
Continuous reactor (free cells) K. pneumoniae Glucose 4.25 [97]
Immobilized cell continuous K. pneumoniae Whey permeate 2.30 [98]
Cell recycle, continuous B. polymyxa Whey permeate 1.04 [95]
Cell recycle, continuous K. pneumoniae Glucose 9.84 [99]
Cell recycle, continuous E. aerogenes Glucose 5.40 [100]
Production of other chemicals in biofilm reactors
Other examples of production of industrial chemicals produced in biofilm reactors include acetic acid or vinegar, lactic acid, succinic acid, and fumaric acid. Acetic acid production in trickling bed biofilm reactors is a mature technology and is exercised at the commercial level [13]. In addition to trickling bed biofilm reactor, a submerged process was also developed in late 1940s. The acetic acid is produced by one of the bacteria grouped in the two genera, Gluconobacter and Acetobacter. The species that are used commercially include Acetobacter aceti, A. pasteurianus, and Gluconobacter oxydans. In the trickling bed biofilm reactor (volume 60 m3), beechwood shavings are packed and the starting material (alcohol solution) is sprayed over the surface. To this solution, initially, nutrients and bacteria are added for growth of biofilm on the beechwood shavings. The liquid trickles to the bottom of the reactor containing acetic acid. In order to increase concentration of acetic acid, the liquid is cooled and pumped back to the top of the reactor. Of the alcohol added, approximately 90% is converted to acetic acid during the trickling process. Approximately 120 gL-1 acetic acid is obtained in 72 h, thus resulting in a productivity of 1.67 gL-1h-1.
Production of lactic acid in biofilm reactors is another example of industrial chemical production in such reactors. Demirci et al. [101] evaluated a number of supports for biofilm formation using lactic acid producing cultures. It has been reported that the best biofilms were obtained with Pseudomonas fragi, Streptomyces viridosporus, and Thermoactinomyces vulgaris when used in combination with polypropylene composite chips. The polypropylene composite chips contained polypropylene and 25% (w/w) agricultural material. The mixture of these components was extruded through an extruder to form chips of desired dimensions. Following this, a number of reports appeared from the same group on synthesizing, evaluating, and using various supports for biofilm formation and lactic acid production [102-105]. In one of the reports [104], lactic acid was produced in repeated batch cultures in a biofilm reactor. The reactor productivity was improved from 2.78 to 4.26 gL-1h-1. A maximum lactic acid concentration of 60 gL-1 was produced in biofilm reactors where plastic composite support was used for adsorption.
In a study on the production of lactic acid by adsorbed cells of Rhizopus oryzae, the culture was immobilized on a fibrous-bed and used in a bioreactor [37]. The fibrous bed was a sheet of 100% cotton cloth onto which the culture was adsorbed. In this reactor (fed-batch), a productivity of 2.5 gL-1h-1 was obtained with a high yield of 90% and a high product concentration of 127 gL-1. Glucose was used as a substrate. When glucose was replaced with cornstarch, yield improved to 100% and productivity decreased to 1.65 gL-1h-1. Using starch as a substrate, a product concentration of 126 gL-1 was achieved.
Other reports on using cell support for cell growth and lactic acid production are those of Park et al. [106] and Sun et al. [107]. Park et al. [106] used 3 gL-1 mineral support (Aid-Plus; ML-50D, Mizusawa Chemical Co., Niigata, Japan) and 5 ppm polyethylene oxide to flocculate the culture and change mycelial morphology from a large pellet to mycelial flocs. Sun et al. [107] immobilized cells of R. oryzae in polyurethane foam cubes. There are other reports on the use of immobilized cell technology to produce lactic acid, however, they have not been mentioned either due to space limitation or studies are not directly related to biofilm formation.
Biofilm reactors have also been used successfully for the production of fumaric acid [108] and mineral ore treatment [13]. In an interesting study, Cao et al. [108] used plastic discs to adsorb cells of R. oryzae to produce fumaric acid from glucose. The use of the biofilm reactor resulted in an increase in reactor productivity from 0.9 gL-1h-1 in a free cell stirred-tank reactor to 4.25 gL-1h-1 in the biofilm reactor. In the latter reactor, fumaric acid concentration up to 85 gL-1 was obtained from 100 gL-1 glucose. The fermentation time was shorter and took 20 h as compared to 72 h in the free cell reactor.
Succinic acid is a chemical that has been produced in biofilm reactors. The industrial potential for succinic acid fermentation was recognized as early as the late 1970s [109]. Succinic acid (HOOCCH2CH2COOH) is a dicarboxylic acid, which can be used as a feedstock chemical for the production of high value products such as 1,4-butanediol, tetrahydrofuran, adipic acid, γ-butyrolactone, and n-methylpyrrolidone [109] for applications in agriculture, food, medicine, plastics, cosmetics, and textiles. In a recent study on succinic acid production using Actinobacillus succinogenes, Urbance et al. [110] employed the customized plastic composite support (PCS) [111] and 20 other different PCS blends with and without mineral salt additions and evaluated 20 simulated repeated-batch fermentations using MgCO3 for pH control and CO2 supply. The customized plastic composite support (PCS) blends were screened for biofilm formation and succinic acid production. Succinic acid concentrations, percentage yield of succinic acid, and biofilm formation for each PCS blend were determined and no correlation between biofilm formation and succinic acid production was observed. However, the customized PCS blend for A. succinogenes in succinic acid production demonstrated 70% yields for succinic acid compared to 64% yield for suspended cell bioreactor [110]. Table 7 shows production of various chemicals in biofilm reactors.
Table 7 Production of various other chemicals in biofilm reactors
Product/Reactor Type Adsorption support Productivity [gL-1h-1] Reference
Acetic acid
Trickling bed biofilm reactor Beechwood shavings 1.67 (120) [13]
Lactic acid
Agitating continuous reactor Fibrous bed (cloth) 2.5 (126) [37]
Fumaric acid
Rotary continuous reactor Plastic discs 4.25 (85) [108]
Strirred-tank (control) None 0.91 [108]
Succinic acid
Repeated batch fermentations Plastic discs - [110]
Numbers in bracket – product concentration in gL-1
Enhanced rates of production of chemicals in biofilm reactors
Length of operation of biofilm reactors
Packed bed reactors often block due to excessive cell growth. It should be noted that reactor blockage depends on a number of factors including cell growth rate, packing density of the support, and supply of nutrients. This type of reactor has been operated ranging from 2 weeks to 3 months. Tyagi and Ghose [4] used a packed bed biofilm reactor of S. cerevisiae for a period of 35 days, while Qureshi et al. [87] used a packed bed biofilm reactor of E. coli for a period of 103 days for ethanol production in continuous operation. However, it was observed that packed bed biofilm reactors of C. acetobutylicum/C. beijerinckii blocked sooner than 103 days due to enhanced cell growth of these cultures. In order to prolong life of the reactor, feed media deficient in nutrients should be attempted as used by Qureshi & Maddox [6] and Qureshi et al. [90]. It has been observed that this type of reactor blocks at the bottom where fresh feed allows excessive cell growth. In the upper part of the reactor, minimal growth occurs due to product inhibition as in case of butanol and ethanol production. In such cases, inverting the reactor can prolong life of reactor. In addition to the reactor blockage due to excessive growth, influent to the reactor plays an important role in prolonging life of the reactor. Reactor feed may contain suspended and particulate solids, in particular with wastewater influents, which may block the reactor. It is suggested that such influents be filtered or centrifuged to remove suspended and particulate solids to prolong reactor's life.
Fluidized bed reactors do not block due to excessive growth and they can be operated for a long period of time (>4 months). It is also viewed that UASB and EGSB can be operated for long periods. Table 8 shows length of operation of different reactors for the production of various chemicals, their productivities and dilution rates. Biofilm reactors are highly productive as compared to other reactor systems. The reader is advised to refer to the production of various chemicals in biofilm reactor systems (in this article) to be able to compare their production rates with the other non-biofilm reactor systems.
Table 8 Length of operation of various biofilm reactors used for the production of different chemicals
Chemical Produced Reactor Type Length of operation [Days] Dilution rate [h-1] (Productivity [gL-1h-1) Reference
Butanol Packed bed 61 days 0.30–1.00 (0.98–4.10) [6]
Fluidized bed >4 months 0.33–1.37 (1.65–5.10) Unpublished data1
Lactic acid Various reactors Reviewed in ref 79 (Table 1) - - [37]
Ethanol Packed bed 35 days 0.12–0.48 (7.80–28.60) [4]
Packed bed 103 days 0.04–0.12 (1.10–2.58) [87]
Packed bed 60 days 0.50–5.76 (5.00–74.88) [86]
1 Qureshi & Maddox
- Not reported
Barriers in biofilm reactors
In adsorbed cell biofilm reactors of C. acetobutylicum, it was identified that there were four different cell types: growing cells, butanol producing cells, dead cells, and inactive cells (non-growing, nutrient requiring) [7]. Cells that were involved in butanol production were only a fraction of the total cells. For example, the concentration of cells in the reactor was approximately 74 gL-1, while the butanol producing cell mass was <10% of the total cells. The amount of dead cells or spores occupied most of the space in the reactor. It is viewed that if sporulation is blocked, the reactor productivity could be increased by many fold. This would improve the process economics of butanol production in biofilm reactors. At this stage we are not aware if this is applicable to the other organisms such as ethanol, 2,3-butanediol, succinic, acetic (vinegar), lactic and fumaric acid producers. It is suggested that this be investigated for the cultures that produce these chemicals. In UASB internal mixing is not optimal which reduces efficiency of the reactor [11]. This produces dead space in the reactor. For that reason expanded granular sludge bed (EGSB) are investigated [11].
Diffusion limitations
Usually biofilms contain multiple layers of cells. The thickness of the biofilm may vary from a few to many μm. An increase in the biofilm particle diameter affects hydrodynamic conditions in the reactor including fluidization characteristics etc [59]. In order to measure the thickness of biofilm in C. acetobutylicum culture (PBR), an electron transmission micrograph was taken of a particle and it was identified that the biofilm was made up of >30 cell layers (Fig. 3c). In order for the cells to be active and be taking part in the reaction, nutrients and substrate must diffuse/penetrate to the inner layers of cells. However, it is likely that the nutrients and substrate are used up by the outer cell layers before they reach the innermost cell layers. If this is true, the innermost layers would neither survive nor take part in the reaction. Another example where the thickness of cell layers is an important consideration is bioparticles in a fluidized bed reactor. In these reactors the size of the bioparticles is much bigger than the bioparticle in PBR and cell layers are >>30. Accumulation of so many cell layers adds to the diffusion resistance to the substrate and nutrients. In order to keep the diffusion resistance to a minimum possible level, the size of the bioparticle should be kept to a minimum level while still keeping productivity of the reactor high. This should increase the rate of reaction and benefit the process economics.
In aerobic biofilm processes, such as oxidative degradation of toxic chemicals and production of acetic acid in trickling bed biofilm reactors, a constant supply of oxygen is essential. The oxygen should be dissolved in the liquid and be transported to the innermost layers. The penetration depth of oxygen should be 100% of the biofilm thickness. If the bioparticle size is large, then the inner layers would be starved of oxygen and the cells would die thus decreasing the conversion efficiency of the process. Supply of oxygen rather than air to the reactor would improve diffusion of oxygen to the inner layers; however, it would add to the cost of the process. Hence, size of the biofilms should also be kept to low to keep the reactor productive. In aerobic wastewater biofilm reactors oxygen is an important substrate/nutrient [12]. For anaerobic systems oxygen is toxic.
In addition to the above limitations, an additional limitation comes from the toxicity of product/s itself. Many of the fermentation products are toxic to the cells that produce them. Examples of such toxic products are those that have been described in the earlier section of this article. Butanol is toxic to the cells of C. acetobutylicum/C. beijerinckii and at higher concentrations it kills the cells. In the biofilm layers, the diffused substrate is converted to the products such as butanol. It is not known how quickly the produced butanol diffuses out of the cell layers. It is conceivable that accumulated butanol or other chemicals kill the cells before it is diffused out. It is also likely that a combination of nutrient deficiency and toxicity affects the cells more adversely.
Industrial/pilot-plant level biofilm reactors
Wastewater treatment
Biofilm reactors have successfully been used in wastewater treatment [9-12,43,59,61]. In these industrial biofilm reactors cell mass concentration as high as 30–40 gL-1 could be maintained [12,58]. As a result of superior efficiency, biofilm reactors are being used throughout the world with a number of full scale application for industrial and wastewater treatment. Examples of these reactors operating in The Netherlands and Brazil are shown in Fig. 6.
Figure 6 Full scale biofilm reactors: (a) biothane biofilm airlift suspension and expanded granular sludge blanket (Biobed) reactors at Gist Brocades, Delft (The Netherlands); (b) Pagues CIRCOX (foreground; 140 m3) and internal circulation (background; 385 m3) reactors at a brewery in Brazil. Reprinted from "Nicolella C, van Loosdrecht MCM, Heijnen SJ: Particle-based biofilm reactor technology. Trends in Biotechnology 2000, 18: 312–320, with permission from Elsevier, United Kingdom.
Acetic acid/vinegar production
Commercial production of acetic acid or vinegar using biofilm reactors has been exercised for many years. Production of these chemicals has been reported by Crueger & Crueger [13]. Large biofilm fermentors of size up to 60,000 L have been used. Often beechwood shavings are used as a support for biofilm formation. For this system, trickling bed reactors have been used with an exit product concentration up to 120 gL-1 and a productivity of 1.67 gL-1h-1. A description of the process has been given in previous sections.
Butanol production
Butanol production in biofilm reactors has been practiced in numerous types of reactors at laboratory scale [6,38,39,44] with superior productivity to batch, fed-batch, and free cell continuous fermentations. Two of the most prominently used reactors are packed bed and fluidized bed reactors. In these reactors, productivities of the order of 4.5–15.8 gL-1h-1 have been achieved as compared to productivities of 0.10–0.38 gL-1h-1 in batch reactors. Given the scenario of increasing petroleum prices, it is suggested that fluidized bed reactors be scaled up to pilot plant level in view to further commercialize this fermentation.
Other processes
Production of other industrial chemicals such as lactic acid and 2,3-butanediol should be exercised at pilot plant level. Nicolella et al. [12] reported that biofilm reactors are in operation at industrial scale throughout the world. Use of biofilm reactors is anticipated to be economical for the production of these industrial chemicals.
Future directions & conclusions
A comparison of biofilm reactors with other reactor systems suggests that biofilm reactors are simple and offer higher productivities than other reactor systems. In biofilm reactors, cells can be adsorbed within the reactor without the use of any chemicals, and the reactors can be operated for long period of times. This would help in reducing the process cost. These reactors are already in use for wastewater treatment and acetic acid/vinegar production by fermentation. It is clear that their use at bench scale has been consistently increasing for the production of various other chemicals. As productivities in these simple biofilm reactors are high, their full potential should be employed for biotechnological/biological conversion processes. For further reading on biofilms and their formation, the reader is referred to the comprehensive articles published by Costerton et al., [112] and O'Toole et al., [113].
Authors' contributions
NQ would like not to mention the contributions made by individual authors. However, it is stated that all the authors made significant contributions to deserve to be contributing authors of this comprehensive article on "Biofilm Reactors."
Note
** Mention of trade names of commercial products in this article is solely for the purpose of providing scientific information and does not imply recommendation or endorsement by the United States Department of Agriculture.
Acknowledgements
N. Qureshi would like to thank Michael A Cotta (United States Department of Agriculture, National Center for Agricultural Utilization Research [USDA/NCAUR], Peoria, IL) for his encouragement during the preparation of this article and reading the manuscript critically. NQ is grateful to Professor Eberhard Morgenroth (Department of Civil and Environmental Engineering, University of Illinois, Urbana, IL) for his help with UASB reactors. Help from Christopher Skory (USDA/NCAUR) is also acknowledged. NQ would like to thank Holly Brining and Mark Maroon for their help during the preparation of this manuscript and Don Fraser for photographic/scanning work. NQ & ISM are grateful to Mr. Doug Hopcroft (AgResearch Ltd., Palmerston North, New Zealand; formerly known as Department of Scientific and Industrial Research [D.S.I.R; Biotechnology Division]) for performing the electron microscopy.
==== Refs
Maddox IS The acetone butanol ethanol fermentation: recent progress in technology Biotechnol Genetic Eng Reviews 1989 7 190 220
Qureshi N Blaschek HP Evaluation of recent advances in butanol fermentation, upstream, and downstream processing Bioproc Biosys Eng 2001 24 219 226 10.1007/s004490100257
Mehaia MA Cheryan M Ethanol production in a hollow fiber bioreactor using Saccharomyces cerevisiae Appl Microbiol Biotechnol 1984 20 100 104 10.1007/BF00252585
Tyagi RD Ghose TK Studies on immobilized Saccharomyces cerevisiae. I. Analysis of continuous rapid ethanol fermentation in immobilized cell reactor Biotechnol Bioeng 1982 24 781 795 10.1002/bit.260240403
Forberg C Haggstrom L Control of cell adhesion and activity during continuous production of acetone and butanol with adsorbed cells Enz Microbial Technol 1985 7 230 234 10.1016/0141-0229(85)90073-0
Qureshi N Maddox IS Continuous solvent production from whey permeate using cells of Clostridium acetobutylicum immobilized by adsorption onto bonechar Enz Microbial Technol 1987 9 668 671 10.1016/0141-0229(87)90125-6
Qureshi N Paterson AHJ Maddox IS Model for continuous production of solvents from whey permeate in a packed bed reactor using cells of Clostridium acetobutylicum immobilized by adsorption onto bonechar Appl Microbiol Biotechnol 1988 29 323 328 10.1007/BF00265814
Demirci A Pometto AL IIIHo KLG Ethanol production by Saccharomyces cerevisiae in biofilm reactors J Ind Microbiol Biotechnol 1997 19 299 304 9439005 10.1038/sj.jim.2900464
Lettinga G van Nelsen AFM Hobma SW de Zeeuw W Klapwijk A Use of the upflow sludge blanket (USB) reactor concept for biological wastewater treatment, especially for anaerobic treatment Biotechnol Bioeng 1980 22 699 734 10.1002/bit.260220402
Lettinga G Roersma R Grin P Anaerobic treatment of raw domestic sewage at ambient temperature using a granular bed UASB reactor Biotechnol Bioeng 1983 25 1701 1723 10.1002/bit.260250703
Seghezzo L Zeeman G van Lier JB Hamelers HVM Lettinga G A review: The anaerobic treatment of sewage in UASB and EGSB reactors Bioresource Technol 1998 65 175 190 10.1016/S0960-8524(98)00046-7
Nicolella C van Loosdrecht MCM Heijnen SJ Particle-based biofilm reactor technology Trends in Biotechnol (TIBTECH) 2000 18 312 320 10.1016/S0167-7799(00)01461-X
Crueger W Crueger C Organic acids "Biotechnology" A textbook of industrial microbiology 1989 Sinauer Associates, Inc., Sunderland. MA
Stoodley P Sauer K Davies DG Costerton JW Biofilms as complex differentiated communities Annu Rev Microbiol 2002 56 187 209 12142477 10.1146/annurev.micro.56.012302.160705
Costerton JW Stewart PS Greenburg EP Bacterial biofilms: a common cause of persistent infections Science 1999 284 1318 1322 10334980 10.1126/science.284.5418.1318
Mortensen KP Conley SN Film fill fouling in counterflow cooling towers: mechanisms and design CTI J 1994 15 10 25
McDonogh R Schaule G Flemming HC The permeability of biofouling layers on membranes J Membr Sci 1994 87 199 217 10.1016/0376-7388(93)E0149-E
Carpentier B Cerf O Biofilm and their consequences with particular reference to hygiene in the food industry J Appl Bacteriol 1993 75 499 511 8294303
Taras M Hakansson K Guieysse B Continuous acetonitrile degradation in packed-bed bioreactor Appl Microbiol Biotechnol 2005 66 567 574 15630519
Hall ER Biofilm reactors in anaerobic wastewater treatment Biotech Adv 1987 5 257 269 10.1016/0734-9750(87)90321-1
Meyer A Wallis FM Development of microbial biofilms on various surfaces for the treatment of heavy metal containing effluents Biotechnol Tech 1997 11 859 863 10.1023/A:1018497813505
Qureshi N Maddox IS Yu PL Novel bioreactors for the ABE fermentation using cells of Clostridium acetobutylicum immobilized by adsorption onto bonechar Fermentation technologies: Industrial applications 1990 London: Elsevier Appl Sci Publ
Maddox IS Qureshi N McQueen J Continuous production of 2,3-butanediol from whey permeate using cells of Klebsiella pneumoniae immobilized on to bonechar New Zealand J Dairy Sci Technol 1988 23 127 132
Sponza DT Anaerobic granule formation and tetrachloroethylene (TCE) removal in an upflow anaerobic sludge blanket (UASB) reactor Enz Microbial Technol 2001 29 417 427 10.1016/S0141-0229(01)00402-1
Annachhatre AP Bhamidimarri SMR Microbial attachment and growth in fixed-film reactors: process startup considerations Biotech Adv 1992 10 69 91 10.1016/0734-9750(92)91352-F
Watnick P Kolter R Biofilm, City of microbes J Bacteriol 2000 182 2675 2679 10781532 10.1128/JB.182.10.2675-2679.2000
Davies DG Chakrabarty AM Geesey GG Exopolysaccharide production in biofilms: substratum activation of alginate gene expression by Pseudomonas aeruginosa Appl Environ Microbiol 1993 59 1181 1186 8476292
Donlan RM Costerton JW Biofilms: survival mechanisms of clinically relevant microorganisms Clin Microbiol Rev 2002 15 167 193 11932229 10.1128/CMR.15.2.167-193.2002
Costerton JW Lewandowski Z Caldwell DE Korber DR Lappin-Scott HM Microbial biofilms Annu Rev Microbiol 1995 49 711 745 8561477 10.1146/annurev.mi.49.100195.003431
Allison DG Ruiz B SanJose C Jaspe C Gilbert P Extracellular products as mediators of the formation and detachment of Pseudomonas fluorescens biofilms FEMS Microbial Lett 1998 167 179 184 10.1016/S0378-1097(98)00386-3
Characklis WG McFeters GA Marshall KC Characklis WG, Marshall KC Physiological ecology in biofilm systems Biofilms 1990 37 New York: John Wiley and Sons 67 72
Qureshi N Karcher P Cotta M Blaschek HP High-productivity continuous biofilm reactor for butanol production Appl Biochem Biotechnol 2004 113–116 713 721 10.1385/ABAB:114:1-3:713
Cowan MM Warren TM Fletcher M Mixed species colonization of solid surfaces in laboratory biofilms Biofouling 1991 3 23 34
Bücks J Mozes N Wandrey C Rouxhet PG Cell adsorption control by culture conditions Appl Microbiol Biotechnol 1988 29 119 128
Donlan RM Biofilms: microbial life on surfaces Emerg Inf Dis 1992 8 881 890
Huang WC Ramey DE Yang ST Continuous production of butanol by Clostridium acetobutylicum immobilized in a fibrous bed bioreactor Appl Biochem Biotechnol 2004 113 887 898 15054240 10.1385/ABAB:115:1-3:0887
Tay A Yang ST Production of L-(+)-lactic acid from glucose and starch by immobilized cells of Rhizopus oryzae in a rotating fibrous bed bioreactor Biotechnol Bioeng 2002 80 1 12 12209781 10.1002/bit.10340
Qureshi N Schripsema J Lienhardt J Blaschek HP Continuous solvent production by Clostridium beijerinckii BA101 immobilized by adsorption onto brick World J Microbiol Biotechnol 2000 16 377 382 10.1023/A:1008984509404
Qureshi N Maddox IS Reactor design for the ABE fermentation using cells of Clostridium acetobutylicum immobilized by adsorption onto bonechar Bioprocess Eng 1988 3 69 72 10.1007/BF00369330
Woolard CR Irvine RL Biological treatment of hypersaline wastewater by a biofilm of halophilic bacteria Water Environ Res 1994 66 230 235
Wobus A Ulrich S Roske I Degradation of chlorophenols by biofilms on semi-permeable membranes in two types of fixed bed reactors Water Sci Technol 1995 32 205 212 10.1016/0273-1223(96)00027-3
Veeresh GS Kumar P Mehrotra I Treatment of phenol and cresols in upflow anaerobic sludge blanket (UASB) process: A review Water Research 2005 39 154 170 15607175 10.1016/j.watres.2004.07.028
Tsuno H Kawamura M Somiya I Anaerobic degradation of pentachlorophenol (PCP) in biological expanded-bed reactor Water Sci Technol 1996 34 335 344 10.1016/0273-1223(96)00663-4
Qureshi N Maddox IS Integration of continuous production and recovery of solvents from whey permeate: use of immobilized cells of Clostridium acetobutylicum in fluidized bed bioreactor coupled with gas stripping Bioprocess Engineering 1991 6 63 69 10.1007/BF00369279
Lazarova V Manem J Bryers JD Innovative biofilm treatment technologies for water and wastewater treatment Biofilm II: Process analysis and applications 2000 New York: Wiley-Liss Press 159 206
Rodgers M Zhan XM Moving-medium biofilm reactors Rev Environ Sci Bio/Technol 2003 2 213 224 10.1023/B:RESB.0000040467.78748.1e
Kargi F Dincer AR Salt inhibition effects in biological treatment of saline wastewater in RBC J Environ Eng 1999 125 966 971 10.1061/(ASCE)0733-9372(1999)125:10(966)
Kargi F Eker S Rotating-perforated-tubes biofilm reactor for high-strength wastewater treatment J Environ Eng 2001 127 959 963 10.1061/(ASCE)0733-9372(2001)127:10(959)
Gönec E Harremoes P Nitrification in rotating disc systems Water Res 1985 19 119 127 10.1016/0043-1354(85)90332-X
Kargi F Uygur A Effect of liquid phase aeration on performance of rotating biodisc contactor treating saline wastewater Environ Technol 1997 18 623 630
Lapara TM Knopka A Nakatsu CH Alleman JE Thermophilic aerobic treatment of a synthetic wastewater in a membrane-coupled bioreactor J Ind Microbiol Biotechnol 2001 26 203 209 11464370 10.1038/sj.jim.7000110
Vogelaar JCT Bouwhuis E Klapwijk A Spanjers H van Lier JB Mesophilic and thermophilic activated sludge post-treatment of paper mill process water Water Res 2002 36 1869 1879 12044086 10.1016/S0043-1354(01)00397-9
Tripathi CS Allen DG Comparison of mesophilic and thermophilic aerobic biological treatment in sequencing batch reactors treating bleached Kraft pulp mill effluent Water Res 1999 33 836 846 10.1016/S0043-1354(98)00260-7
Quesnel D Nakhla G Utilization of an activated sludge for the improvement of an existing thermophilic wastewater treatment system Environ Eng 2005 131 570 578 10.1061/(ASCE)0733-9372(2005)131:4(570)
Jeris JS Owen RW Hickey R Flood F Biological fluidized-bed treatment for BOD and nitrogen removal J Water Pollut Control Fed 1977 49 816 831
Rabah FKJ Dahab MF Nitrate removal characteristics of high performance fluidized-bed biofilm reactors Water Res 2004 38 3719 3728 15350424 10.1016/j.watres.2004.07.002
Jesis JS Owen RW Biological fluid-bed treatment for BOD and nitrogen removal J Water Pollut Control Fed 1977 49 816 821
Shieh WK Keenan JD Fluidized bed biofilm reactor for wastewater treatment Adv Biochem Eng Biotechnol 1986 33 131 169
Iza J Fluidized bed reactors for anaerobic wastewater treatment Water Sci Technol 1991 24 109 132
Ye FX Shen DS Feng XS Anaerobic granule development for removal of pentachlorophenol in an upflow anaerobic sludge blanket (UASB) reactor Process Biochem 2004 39 1249 1256 10.1016/S0032-9592(03)00247-4
Schmidt JE Batstone DJ Angelidaki I Improved nitrogen removal in upflow anaerobic sludge blanket (UASB) reactors by incorporation of Annammox bacteria into the granular sludge Water Sci Technol 2004 49 69 76 15303725
Horber C Christiansen N Arvin E Ahring BK Improved dechlorinating performance of upflow anaerobic sludge blanket reactors by incorporation of Dehalospirillum multivorans into granular sludge Appl Env Microbiol 1998 64 1860 1863 9572963
Iza J Keenan PJ Switzenbaum MS Anaerobic treatment of municipal solid waste landfill leachate: Operation of a pilot scale hybrid UASB/AF reactor Water Sci Technol 1992 25 255 264
Fang HHP Hui HH Effect of heavy metals on the methanogenic activity of starch-degrading granules Biotechnol Lett 1994 16 1091 1096 10.1007/BF01022409
Satoh H Nakamura Y Ono H Okabe S Effect of oxygen concentration on nitrification and denitrification in single activated sludge flocs Biotechnol Bioeng 2003 83 604 607 12827702 10.1002/bit.10717
Pochana K Keller J Study of factors affecting simultaneous nitrification and denitrification (SND) Water Sci Technol 1999 39 61 68 10.1016/S0273-1223(99)00123-7
Robertson LA van Niel EDJ Torremans RAM Kuenen JG Simultaneous nitrification and denitrification in aerobic chemostat cultures of Thiosphaera pantotropha Appl Environ Microbiol 1988 54 2812 2818 16347780
Robertson LA Cornelisse R De Vos P Hadioetomo R Kuenen JG Aerobic denitrification in various heterotrophic nitrifiers Antonie van Leeuwenhoek (Historical Archive) 1989 56 289 299 10.1007/BF00443743
Jensen K Sloth NP Risgaard-Petersen N Rysgaard S Revsbech NP Estimation of nitrification and denitrification from microprofiles of oxygen and nitrate in model sediment systems Appl Environ Microbiol 1994 60 2094 2100 16349295
van Niel EDJ Braber KJ Robertson LA Kuenen JG Heterotrophic nitrification and aerobic denitrification in Alcaligenes faecalis strain TUD Antonie van Leeuwenhoek (Historical Archive) 1992 62 231 237 10.1007/BF00582584
Robertson LA Kuenen JG Thiosphaera pantotropha gen nov sp nov, a facultative anaerobic, facultative autotrophic sulphur bacterium J Gen Microbiol 1983 129 2847 2855
Robertson LA Kuenen JG Aerobic denitrification: a controversy revived Archive Microbiol 1984 139 351 354 10.1007/BF00408378
Brindle K Stephenson T The application of membrane biological reactors for the treatment of wastewaters Biotechnol Bioeng 1996 49 601 610 10.1002/(SICI)1097-0290(19960320)49:6<601::AID-BIT1>3.0.CO;2-S
Casey E Glennon B Hamer G Oxygen mass transfer characteristics in a membrane-aerated biofilm reactor Biotechnol Bioeng 1999 62 183 192 10099528 10.1002/(SICI)1097-0290(19990120)62:2<183::AID-BIT8>3.0.CO;2-L
Debus O Wanner O Degradation of xylene by a biofilm growing on a gas permeable membrane Water Sci Technol 1992 26 607 616
Brindle K Stephenson T Nitrification in a bubbleless oxygen mass transfer membrane bioreactor Water Sci Technol 1996 34 261 267 10.1016/S0273-1223(96)00813-X
Kennes C Thalasso F Waste Gas Biotreatment Technology J Chem Technol Biotechnol 1998 72 303 319 10.1002/(SICI)1097-4660(199808)72:4<303::AID-JCTB903>3.3.CO;2-P
Ottengraf SPP Biological systems for waste gas elimination Trends Biotechnol 1987 5 132 136 10.1016/0167-7799(87)90007-2
Koe LCC Yang F A bioscrubber for hydrogen sulfide removal Water Sci Technol 2000 41 141 145
Ottengraf S Van Den Oever A Kinetics of organic compound removal from waste gases with a biological filter Biotechnol Bioeng 1983 25 3089 3102 10.1002/bit.260251222
Shinabe K Oketani S Ochi T Matsumura M Characteristics of hydrogen sulfide removal by Thiobacillus thiooxidans KS1 isolated from a carrier-packed biological deodorization system J Ferment Bioeng 1995 80 592 598 10.1016/0922-338X(96)87737-3
De Beer D Stoodley P Roe F Lewandowski Z Effects of biofilm structures on oxygen distribution and mass transport Biotechnol Bioeng 1994 43 1131 1138 10.1002/bit.260431118
Morgenroth E Schroeder ED Chang DPY Scow KM Nutrient limitation in a compost biofilter degrading hexane J Air Waste Managem Assoc 1996 46 300 308
Bland RR Chen HC Jewell WJ Bellamy WD Zall RR Continuous high rate production of ethanol by Zymomonas mobilis in an attached film expanded bed fermentor Biotechnol Lett 1982 4 323 328 10.1007/BF00132834
Krug TA Daugulis AJ Ethanol production using Zymomonas mobilis immobilized on an ion exchange resin Biotechnol Lett 1983 5 159 164 10.1007/BF00131895
Kunduru MR Pometto AL III Continuous ethanol production by Zymomonas mobilis and Saccharomyces cerevisiae in biofilm reactors J Ind Microbiol Biotechnol 1996 16 249 256
Qureshi N Brining H Iten L Dien B Nichols N Saha B Cotta MA Adsorbed cell dynamic biofilm reactor for ethanol production from xylose and corn fiber hydrolysate The 36th Great Lakes Regional Meeting of the American Chemical Society, Peoria, IL, October 17–20, 2004
Rogers PL Lee KJ Skotnicki ML Tribe DE Ethanol production by Zymomonas mobilis Adv Biochemical Eng 1982 23 37 84
Welsh FW Williams RE Veliky IA Solid carriers for a Clostridium acetobutylicum that produced acetone and butanol Enz Microbial Technol 1987 9 500 502 10.1016/0141-0229(87)90105-0
Qureshi N Lai LL Blaschek HP Scale-up of a high productivity continuous biofilm reactor to produce butanol by adsorbed cells of Clostridium beijerinckii BA101 Transactions Institution Chemical Engineers (Trans IChemE) (Chemical Engineering Research and Design) 2004 82 164 173
Qureshi N Blaschek HP Shetty K, Pometto A, Paliyath G Butanol production from agricultural biomass Advances in Food Biotechnology 2005 New York, NY: CRC Press
Ezeji TC Qureshi N Blaschek HP Durrie P Industrially relevant fermentations Handbook on Clostridia 2005 New York, NY: CRC Press
Pierrot P Fick M Engasser JM Continuous acetone butanol fermentation with high productivity by cell ultrafiltration and recycling Biotechnol Lett 1986 8 253 256 10.1007/BF01030507
Afschar AS Biebl H Schaller K Schugerl K Production of acetone and butanol by Clostridium acetobutylicum in continuous culture with cell recycle Appl Microbiol Biotechnol 1985 22 394 398 10.1007/BF00252779
Shazer WH Speckman RA Continuous fermentation of whey permeate by Bacillus polymyxa J Dairy Sci 1984 67 1216 1218
Syblayrolles JM Goma G Butanediol production by Aerobactor aerogenes NRRL B199: effect of initial substrate concentration and aeration agitation Biotechnol Bioeng 1984 26 148 155 10.1002/bit.260260207
Ramachandran KB Goma G Effect of oxygen supply and dilution rate on the production of 2,3-butanediol in continuous bioreactor by Klebsiella pneumoniae Enz Microbial Technol 1987 9 107 111 10.1016/0141-0229(87)90152-9
Lee HK Maddox IS Continuous production of 2,3-butanediol from whey permeate using Klebsiella pneumoniae immobilized in calcium alginate Enz Microbial Technol 1986 8 409 411 10.1016/0141-0229(86)90147-X
Ramachandran KB Goma G 2,3-butanediol production from glucose by Klebsiella pneumoniae in a cell recycle system J Biotechnol 1988 9 39 46 10.1016/0168-1656(88)90013-2
Zeng AP Biebl H Deckwer WD Production of 2,3-butanediol in a membrane bioreactor with cell recycle Appl Microbiol Biotechnol 1991 34 463 468 10.1007/BF00180571
Demirci A Pometto AL IIIJohnson KE Evaluation of biofilm reactor solid support for mixed-culture lactic acid production Appl Microbiol Biotechnol 1993 38 728 733 10.1007/BF00167135
Demirci A Pometto AL IIIJohnson KE Lactic acid production in a mixed-culture biofilm reactor Appl Environ Microbiol 1993 59 203 207 16348843
Ho KLG Pometto AL IIIHinz PN Demirci A Nutrient leaching and end product accumulation in plastic composite supports for L-(+)-lactic acid biofilm fermentation Appl Environ Microbiol 1997 63 2524 2532 16535636
Ho KLG Pometto AL IIIHinz PN Optimization of L-(+)-lactic acid production by ring and disc plastic composite supports through repeated-batch biofilm fermentation Appl Environ Microbiol 1997 63 2533 2542 9212403
Ho KLG Pometto AL IIIHinz PN Dickson JS Demirci A Ingredient selection for plastic composite supports for L-(+)-lactic acid biofilm fermentation by Lactobacillus casei subsp. rhmnosus Appl Environ Microbiol 1997 63 2516 2513 9212402
Park EY Kosakai Y Okabe M Efficient production of L-(+)-lactic acid using mycelial cotton-like flocs of Rhizopus oryzae in an air-lift bioreactor Biotechnol Prog 1998 14 699 704 9758658 10.1021/bp9800642
Sun Y Li YL Bai S Modeling of continuous L(+)-lactic acid production with immobilized R. oryzae in an airlift bioreactor Biochem Eng J 1999 3 87 90 10.1016/S1369-703X(98)00044-8
Cao N Du J Gong CS Tsao GT Simultaneous production and recovery of fumaric acid from immobilized Rhizopus oryzae with a rotary biofilm contactor and an adsorption column Appl Env Microbiol 1996 62 2926 2931 16535381
Zeikus JG Jain MK Elankovan P Biotechnology of succinic acid production and markets for industrial products Appl Microbiol Biotechnol 1999 51 545 552 10.1007/s002530051431
Urbance SE Pometto AL IIIDiSpirito AA Demirci A Medium evaluation and plastic composite support ingredient selection for biofilm formation and succinic acid production by Actinobacillus succinogenes Food Biotechnol 2003 17 53 65 10.1081/FBT-120019984
Cotton JC Pometto AL IIIGvozdenovic-Jeremic J Continuous lactic acid fermentation using a plastic composite support biofilm reactor Appl Microbiol Biotechnol 2001 57 626 630 11778870 10.1007/s002530100820
Costerton JW Cheng KJ Geesey GG Ladd TI Nickel JC Dasgupta M Marrie TJ Bacterial biofilms in nature and disease Annu Rev Microbiol 1987 41 435 464 3318676 10.1146/annurev.mi.41.100187.002251
O'Toole G Kaplan HB Kolter R Biofilm formation as microbial development Annu Rev Microbiol 2000 54 49 79 11018124 10.1146/annurev.micro.54.1.49
|
16122390
|
PMC1236956
|
CC BY
|
2021-01-04 16:24:35
|
no
|
Microb Cell Fact. 2005 Aug 25; 4:24
|
utf-8
|
Microb Cell Fact
| 2,005 |
10.1186/1475-2859-4-24
|
oa_comm
|
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-381611150410.1186/1477-7827-3-38ResearchEvaluation of the true precocious puberty rats induced by neonatal administration of Danazol: Therapeutic effects of nourishing "Yin"- removing "Fire" Chinese herb mixture Tian Zhanzhuang [email protected] Hong [email protected] Yan [email protected] Depei [email protected] Boying [email protected] Department of Neurobiology and Integrative Medicine, Shanghai Medical College of Fudan University (Formerly Shanghai Medical University), P.O. Box 291, 138 Yi-Xue-Yuan Road, 200032 Shanghai, P. R. China2 Children's Hospital of Fudan University, 183 Feng-Lin Road, 200032 Shanghai, P. R. China2005 22 8 2005 3 38 38 1 7 2005 22 8 2005 Copyright © 2005 Tian et al; licensee BioMed Central Ltd.2005Tian 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
Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture, a traditional herb-based formulation, has been successfully used for the management of idiopathic true precocious puberty (IPP) for more than thirty years. Precocious puberty rat model by neonatal administration of Danazol was used to investigate the effects of the herb mixture on the advanced sexual development of the rats, and the expression of hypothalamic gonadotropin-releasing hormone (GnRH), which is the important regulator for the hypothalamus-pituitary-gonadal axis, particularly at puberty.
Methods
Female Sprague-Dawley rats were divided into five groups: intact normal (N), IPP model (M), vehicle with no IPP (V), IPP model exposed to herb mixture (HM) and IPP model exposed to saline (S). Rats at 5 days of age were given a single subcutaneous injection of 300 microgram of Danazol dissolved in 25 microliter vehicle of propylene glycol-ethanol (1:1, v/v), to establish the precocious puberty model. From the day 15, rats in HM and S groups were continuously fed with either Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture 2 ml or saline 2 ml, until 3 consecutive regular estrous cycles were established. The day of vaginal opening and the day of setup regular estrous cycle of the rats were observed. Blood concentration of estrogen was determined by radioimmunoassay. Immunohistochemistry and RT-PCR analysis were used to explore the expression of GnRH.
Results
The day of vaginal opening and first estrous showed significant advancement in M compared with N and V (p < 0.05, respectively). The blood estrogen level increased significantly in M compared with those in other groups (about 28 days of age, at the time of vaginal opening in M rats) (p < 0.05, respectively). GnRH cells in rostral medial septum (MS), Broca diagonal band nucleus (DBB) and the medial preoptic area (MPOA), were calculated. The number in M was less than those in N and V (p < 0.05, respectively). The number was significantly higher in HM than that in M (p < 0.05). The GnRH mRNA expression increased significantly in M compared with that in N and V (p < 0.05).
Conclusion
The true precocious puberty model by neonatal administration of Danazol in female rats showed augmented expression of hypothalamic GnRH; the Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture down-regulated the increased GnRH expression, and significantly delayed the sexual development of the precocious puberty rat.
==== Body
Background
Sexual precocity is one of the most popular endocrine disorders in children, incidence of which is about 0.6% throughout the world [1]. It is 10 times more common in girls than in boys of the disease. Precocious puberty is the appearance of the secondary sexual characteristics before age 8 years and can be mainly classified into true precocious puberty (GnRH-dependent sexual precocity) and incomplete isosexual precocity (GnRH-independent sexual precocity). The term of true precocious puberty properly applies only to sexual precocity mediated by premature activation of the hypothalamic-pituitary-ovarian axis usually before 8 year-old. Though central nervous system tumors such as hamartomas and astrocytomas may cause true precocious puberty, most cases have no organic disease, which is defined idiopathic true precocious puberty (IPP) [2].
Though experimental models of precocious puberty have been induced in female rats by neonatal injection of testosterone, estradiol or melatonin, these rats developed persistent vaginal estrous or disturbance of cyclicity (predominance of estrous smear) shortly after the day of first estrous. It has been reported that the neonatal administration of Danazol may affect the hypothalamic pituitary axis with the rapid rate of maturation, which may serve as a model for analyzing true precocious puberty [3].
Several medications have been reported to be effective against true precocious puberty, which include the GnRH analogues [4], progesterone prescriptions and Chinese herbal medicine (CHM). Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture, a traditional herb-based formulation, has been successfully used for the management of IPP by us for more than thirty years. It has been clinically verified to successfully modulate the course of pubertal development and optimize skeletal development in children with precocious puberty, but without notable side effects [5]. The present study was to observe the effects of the herb mixture on the true precocious rat model.
Methods
Animals
Female Sprague-Dawley rats at 3 days of age in company with the maters were purchased from Medical Experimental Animals Center of Fudan University (Shanghai, China). Animals were housed under laminar flow in an isolated room with controlled temperature and at a 12 /12 (light /dark) schedule. The model [3] litters at day 5 (the day of birth was termed day 1) were given a single subcutaneous injection of 300 μg of Danazol (Hualian Pharm Ltd, Shanghai, China) dissolved in 25 μl vehicle of propylene glycol-ethanol (1:1, v/v), and allowed to grow without further treatment. The animals were weaned on day 23, and were examined daily for vaginal opening afterwards, and then the daily vaginal smears were examined. All experimental procedures involving the use of animals were conducted in accordance with NIH Guidelines and were reviewed and approved by the Animal Use and Care Committee for the Fudan University.
In the Experiment 1, rats were used to observe the day of vaginal opening and the day of setup regular oestrous cycle, which were divided into five groups: intact normal (N), IPP model (M), vehicle with no IPP (V), IPP model exposed to herb mixture (HM) and IPP model exposed to saline (S). Daily vaginal smears were checked until 3 consecutive regular 4 or 5 days estrous cycles were established. The day of diestrous I of the first cycle in the 3 consecutive regular estrous cycles was determined as the day of setup regular estrous cycle. From the day 15, rats in HM and S groups were continuously fed with either Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture 2 ml or saline 2 ml, until 3 consecutive regular estrous cycles were established.
RIA of blood estrogen concentration
At the time of vaginal opening in model rats, the blood samples of all the rats were collected from tail veins respectively. The plasma was separated by centrifugation and stored at -80°C until assayed. Concentration of E2 was determined by a double-antibody RIA using the kit purchased from the Shanghai Institute of Biological Products (Shanghai, China.). The sensitivity of the kit was 1.4 pg/ml, and the intra- and inter-assay coefficients of variation were 3.7–8.0% and 4.74–7.7%.
Immunohistochemistry analysis
In the Experiment 2, thirty rats divided into the same five groups were used to investigate the expression of GnRH in the rostral medial septum (MS), Broca diagonal band nucleus (DBB) and the medial preoptic area (MPOA) of the rats by immunohistochemistry according to the Atlas [6]. At the days of vaginal opening in the model group, following anesthesia, all of the animals were exsanguinated with normal saline. Perfusion done, the brains were removed and split equally along the third ventricle. One half of the brain was postfixed for >48 h in 4% paraformaldehyde in 0.1 M PB (PH 7.4) with 30% sucrose, and their sections were sliced at 35 μm thickness on a vibratome microslicer and stored at 4°C in tissue culture wells containing 0.1 M PBS (PH 7.4) plus 0.02% sodium azide until further processed. The other half of the brain was snap-frozen in liquid nitrogen, and then stored at -80°C.
Washed in PB for 30 min at room temperature (RT), the floating sections were incubated in PB containing 10% bovine serum albumin (BSA) for 2 h at RT, then with both rabbit anti-GnRH (1:2000, Chemicon International Inc.) diluted in PB containing 1% BSA, 0.02% sodium azide, and 0.4% Triton-X 100 at 37°C for 2 h, and then at 4°C for 70 h. When they had been rinsed in PB (× 3, 5 min each), the sections were incubated in secondary antibody solutions (goat anti-rabbit IgG conjugated to peroxidase 1:200, Sino-American Technology Company, China) with the same blocking sera for 2 h at RT. After that, they were washed (× 3, 30 min), and diaminobenzidine (DAB) was used as chromogen (Vectastain Elite kits, Vector Labs).
Tissue collection andtotal RNA preparation
The other half of the brain in Experiment 2 was used to investigate the expression of GnRH mRNA by RT-PCR. The target regions, including mediobasal hypothalamus and the suprachiasmatic-preoptic area were dissected (limited anteriorly by the optic chiasma, laterally by the hypothalamic fissures, posteriorly by the mammilary bodies and in depth by the subthalamic sulcus). Total hypothalamic RNA was extracted using 'TRIzol Regent' (Biobasic Inc., Canada) according to the manufacturer's instructions. The purity and integrity of the RNA were checked spectroscopically and by gel electrophoresis before carrying out the analytical procedures.
RNA analysis
Tissue RNA (2 μg) was reverse transcribed, in a final volume of 20 μl, using 200 IU M-MLV reverse transcriptase in the presence of 25 pmol GnRH specific downstream primer (Sangon Inc), 0.5 mM deoxy-NTP and 20IU Rnasin (from Promega) for 60 min at 42°C, then heat denatured for 5 min at 95°C. 5 μl cDNAs were further amplified by PCR using 25 pmol of upstream primer (Sangon Inc) for GnRH: 5'-ATT CTA CTG ACT TGG TGC GTG-3'; downstream, 5'-GGA ATA TGT GCA ACT TGG TGT-3' [7]. PCR was performed for 30 cycles (1 min at 94°C, 45 sec at 62°C, 1 min at 72°C) in the presence of Taq DNA polymerase (3U per tube) and 2.2 mM magnesium chloride (from Promega) in a final volume of 50 μl. To check the presence of DNA contamination RT-PCR was performed on 2 μg of total RNA without M-MLV reverse transcriptase (negative control). An internal control (β-actin, upstream, 5'-AAG CAG GAG TAT GAC GAG TCC G-3'; downstream, 5'-GCC TTC ATA CAT CTC AAG TTG G-3') for each RT-PCR was performed to account for procedural variations. For each sample 5 μl of PCR amplification products were analyzed on 2% agarose gels and stained with ethidium bromide. Standard DNA (100 bp DNA ladder Promega) was run to provide the appropriate size marker. The RT-PCR products were extracted and purified from agarose gel by Golden Beads Gel Extraction kit (Sangon Inc., China) and sequenced using radioactive dideoxychain terminating method (Sangon Inc., China). The intensities of the bands were evaluated by Image Master Software (SYDR-1990, SYNGENE, U.S.A.).
Statistical analysis
All data are presented as means ± S.E.M. Statistical analysis was performed on raw data using one-way analysis of variance (ANOVA), with the significance concentrations of p < 0.05 in two-tailed testing chosen. Comparisons among groups were made using the Student's t-test.
Results
The day of vaginal opening and regular estrous cycle of the rats
The day of vaginal opening and regular estrous cycle were significantly (p < 0.01, respectively) advanced in M than that in N and V, and there was no difference between N and V groups (Table 1). The day of vaginal opening and regular estrous cycle in HM were significantly (p < 0.01) delayed than those in M and S (Table 1).
Table 1 The day of vaginal opening and regular estrous cycle of the rats
N (n = 18) V (n = 18) M (n = 18) HM (n = 18) S (n = 18)
Day of vaginal opening 34.00 ± 0.98 34.33 ± 1.09 25.97 ± 2.24* 32.56 ± 1.04# 24.78 ± 1.01*
Day of regular estrous cycle 41.26 ± 2.35 45.00 ± 1.19 32.43 ± 0.75* 40.58 ± 0.84# 34.56 ± 0.89*
Blood estrogen (pg/ml) 22.64 ± 3.69 24.58 ± 3.87 50.36 ± 5.54* 30.98 ± 7.02# 51.14 ± 4.98*
*p < 0.05 vs N and V, respectively; # p < 0.05 vs M and S, respectively
N: intact normal, M: IPP model, V: vehicle with no IPP, HM: IPP model exposed to herb mixture, S: IPP model exposed to saline
Blood concentration of estrogen
The blood E2 concentration in M is twofold higher than that in N (p < 0.05). While the E2 level was lower in HM than in S (p < 0.05). There were no differences between N and V, M and S (Table 1).
Effects of herb mixture on GnRH expression by immunohistochemistry
GnRH cells bodies, which were abundant in MS, DBB and MPOA, were calculated. The GnRH cells number in M was less than that in V (p < 0.05), and there was no difference between N and V (Fig 1). The number was significantly higher in HM than that in S (p < 0.05), and no statistical difference was observed between HM and N (Fig 1). There was no significant difference between M and S (Fig 1). The tissue sections processed for immunohistochemistry using antiserum after preabsorption with excessive antigens and omission of primary antibody showed the staining as expected.
Figure 1 Effects of herb mixture on GnRH expression by immunohistochemistry. N: intact normal, M: IPP model, V: vehicle with no IPP, HM: IPP model exposed to herb mixture, S: IPP model exposed to saline. A: The statistical analysis of (spiny GnRH cells/total cells) % (n = 6). * p < 0.05 vs M; ^ P < 0.05 vs HM. B: Total GnRH cells of the rats (n = 6). * p < 0.05 vs N and V, respectively; ^ P < 0.05 vs HM. C: Light micrography illustrating the different types of GnRH cells at the region of MS at low power (n = 6). Sm: smooth type cell; Sp: spiny type cell.
The subtypes of all the observed GnRH neurons were discerned under high microscope (× 400). The classification of spiny type and smooth type GnRH neurons was according to the report by Witkin [8] (Fig 1). The percent of spiny type GnRH neurons was significantly higher in M than those in N and HM (p < 0.05, respectively) (Fig 1). There were no differences between N and V, M and S.
Effects of herb mixture on GnRH mRNA expression by RT-PCR
Comparison of the amplified PCR fragment with rat GnRH sequence revealed 100% homology (data not shown). Densitometric analysis of the mRNA concentration using GnRH/β-actin expressed as the mean with SEM. The ratio of GnRH to β-actin in the M increased significantly compared with that in the N and V (p < 0.05), and the ratio in the HM decreased, with no difference showed between HM and N. (Fig 2). There was no significant difference between M and S.
Figure 2 Effects of herb mixture on GnRH mRNA expression by RT-PCR. The upper picture shows the gel electrophoresis of the RT-PCR products for the GnRH. Densitometric analysis of the mRNA concentration using GnRH/β-actin expressed as the mean with SEM bar (n = 6) in each column indicated in the lower panel. N: intact normal, M: IPP model, V: vehicle with no IPP, HM: IPP model exposed to herb mixture, S: IPP model exposed to saline. *p < 0.05 vs N and V, respectively; ^ p < 0.05 vs M and S, respectively.
Discussion
Mammalian sexual maturation and adult reproductive function are centrally controlled by a subset of neuroendocrine neurons that produce GnRH. Recent ten years study indicated, GnRH neurons are the final common pathway for the neuronal control of gonadotropin secretion, which will in turn stimulate gametogenesis and gonadal hormone secretion [9]. In humans and rats, the onset of puberty is highlighted by the augmentation of pulsatile GnRH secretion [10,11]. GnRH release is in a state of low level before puberty, an increased pulsatile release of GnRH is essential for the onset of puberty. True precocious puberty is commonly idiopathicly mediated by premature increasing secretion of GnRH, and is readily treated with potent GnRH agonists [4].
Though GnRH neurons in CNS are paucity, about 600–2,000 cells in mammals, the highest density of the cells is located in MS, DBB and MPOA, approximately 50 to 70% of which project to the median eminence (ME), thus contributing to the control of LH secretion from the adenohypophysis [12]. From this point of view, we selected these regions to observe the change of GnRH expression. Morishito et al has reported the animal model that neonatal treatment with Danazol may induce the true precocious puberty in female rats [3], but the hypothalamic GnRH expression was not examined. In the present study, the high level of GnRH mRNA expression was observed in the M, but the number of the GnRH neuron decreased. The possible mechanism of the inconsistency change might due to the super-release of hypothalamic GnRH in the model rats. Nevertheless, our results, the GnRH mRNA expression showed a higher level in the precocious puberty model rats than in the same age normal rats, the blood estrogen concentration in the model group is twofold higher than that in normal one, along with the day of vaginal opening and the onset of regular estrous cycle advanced significantly in the model rats, give more evidence to sustain the hypothesis that the model induced by neonatal treatment with Danazol is a true precocious puberty model, and might be served to study the onset mechanism of puberty. Besides, though experimental models of precocious puberty have been induced in female rats from 15 to 30 days of age [11,13] or during neonate, by testosterone, estradiol, melatonin and so on, this model appears the subsequent normal ovulatory cycles during adulthood [3].
Danazol, an isoxazol derivative of 17α-ethinylestosterone, is known to have various effects on the reproductive system. Danazol binds to androgen and progesterone receptors, and possesses weak androgenic activity in rats. But the mechanism and the action site of Danazol, administered during the neonatal period, induces a true precocious puberty is not clear. It is generally accepted that the rat hypothalamus is immature at birth and that manipulations of maturation process are possible between days 1 and 10 of life. From these viewpoints, it is possible that the neonatal administration of Danazol may affect the hypothalamic pituitary axis with the rapid rate of maturation, producing a true precocious puberty.
GnRH neurons display a range of shapes from smooth-contoured to extremely irregular or spiny, which reflects or can be modified by the gonadal steroid milieu [8]. The spiny on the GnRH neuron has been identified the synapse by its ultrastructure, suggesting that more synapse connections exist among GnRH neurons and other components in the GnRH network. In the present results, the percent of spiny type GnRH neuron was significantly higher in M than those in N and HM, which might indicate that, on one hand, the gonadal steroid concentration in the local brain region was high; on the other hand, more factors were involved in the process. The local brain aromatization has been observed to be increased in the M [14], which indirectly reflects the increased estrogen concentration. The factors controlling the increased and abnormal secretion of GnRH at the onset of precocious puberty are now being investigated in the same model by us, in which TGF might be an important factor. But further study is carried out. The animals in M and HM groups displayed neural changes to the different degree, suggesting that the herb mixture played certain regulatory effects on the precocious puberty rats.
We have, successfully used the Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture, for the management of IPP for more than thirty years. In 1990', one hundred and six clinical cases have been reported [5]. Before and after treatment, GnRH stimulating test, size of uterus and ovary, X-ray bone age measurement were adopted to estimate the effect of the herb mixture on HPOA and development of internal genitalia and skeleton. Clinical studies indicate that the herb mixture could markedly reduce the blood levels of FSH, LH, E2 of the patients, and the volume of uterus and ovary as well. The herb mixture could also inhibit the excessive functional activities of osteoblasts and the decelerated linear growth of skeleton. Besides, the cheap-effective therapeutic approach is more suitable to our national condition than the high-price GnRH agonist.
Though the physiology of human and rodent is not same, the therapeutic effects of the herb mixture on not only true precocious puberty patient but also rat have been observed. The mechanism of the effective drug has been further studied focusing on the neuroendocrine gene expression with modern medical techniques. Recently, seven differential displayed genes between M and HM rats in the same target brain region, have been screened out, by the means of RNA arbitrarily primed PCR (RAP-PCR) (unpublished data). Hopefully, with the all-right therapeutic effect, the herb mixture compound might bring more profits for patients throughout the world.
Conclusion
The true precocious puberty model by neonatal administration of Danazol in female rats showed augmented expression of hypothalamic GnRH; the Nourishing "Yin"-Removing "Fire" Chinese Herb Mixture down-regulated the increased GnRH expression, and significantly delayed the sexual development of the precocious puberty rat.
Authors' contributions
Zhanzhuang Tian and Hong Zhao designed the study, performed the animal and molecular genetic studies, and drafted the manuscript. Yan Sun performed the statistical analysis. Depei Cai participated in the coordination. Boying Chen conceived of the study, and participated in its design and coordination. All authors read and approved the final manuscript.
Acknowledgements
This project was financed by the Younger Scientific Fund of Fudan University (2004) and Shanghai Medical College (2004).
==== Refs
Ge QS Precocious puberty J Reprod Med 2000 9 377 380
Samuel YenSC Reproductive Endocrinology: physiology, pathophysiology and clinical management 2001 4 W.B. Saunders Company 388 412
Morishito H Takemoto M Kondo H Higuchi K Aono T Induction of true precocious puberty by neonatal treatment with danazol in female rats Neurosci Lett 1993 157 33 36 8233026 10.1016/0304-3940(93)90636-Y
Heger S Sippell WG Partsch CJ Gonadotropin-releasing hormone analogue treatment for precocious puberty. Twenty years of experience Endocr Dev 2005 8 94 125 15722620
Cai D Zhang W Regulative actions of the Chinese drugs for tonifying the kidney on gene expression of the hypothalamic GnRH, pituitary FSH, LH and osteoblastic BGP J Tradit Chin Med 2005 25 58 61 15889526
Bao XM Shu SY Stereostic Brain Atlas 1991 Beijing People's Health Press 22 30
Tian ZZ Zhao H Chen BY Expression of GnRH and GnRH receptor mRNA in female rats with danazol-induced precocious puberty Chin J Endocrinol Metab 2003 19 399 401
Witkin J Morphology of luteinizing hormone-releasing hormone neurons as a function of age and hormonal condition in the male rat Neuroendocrinology 1989 49 344 348 2654690
Terasawa EI Fernandez DL Neurobiological mechanism of the onset of puberty in primates Endocrin Rev 2001 22 111 150 10.1210/er.22.1.111
Clive MD Jonathan R Timothy WJ Robert PM Prepubertal increase in gonadotropin-releasing hormone mRNA, gonadotropin-releasing hormone precursor, and subsequent maturation of precursor processing in rats J Clin Invest 1992 90 2496 2501 1469100
Pinyerd B Zipf WB Puberty-timing is everything! J Pediatr Nurs 2005 20 75 82 15815567 10.1016/j.pedn.2004.12.011
Zhao H Tian ZZ Chen BY Progress on GnRH cells model (review) Prog Physiol 2004 35 185 187 (in Chinese)
Pinilla L Lopez F Gonzalez D Fernandez C Criado JES Aguilar E Luteinizing-hormone-mediated precocious puberty induced in female rats by a prepubertal pituitary graft Neuroendocrinology 1989 50 495 499 2514388
Tian ZZ Zhao H Chen BY Decreased hypothalamic aromatase in female rats of precocious puberty Neurosci Lett 2004 366 92 96 15265597 10.1016/j.neulet.2004.05.012
|
16111504
|
PMC1236957
|
CC BY
|
2021-01-04 16:37:12
|
no
|
Reprod Biol Endocrinol. 2005 Aug 22; 3:38
|
utf-8
|
Reprod Biol Endocrinol
| 2,005 |
10.1186/1477-7827-3-38
|
oa_comm
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.